Skip to content

DPX — Settlement & Intelligence for AI Agents MCP server

io.github.untitledfinancial/dpxuntitledfinancialstreamable-httpv2.6.0Scanned

Compliance-gated stablecoin settlement + x402 macro/ESG/climate intelligence for agents. 115 tools.

113 tools: 5 can delete or cancel things, 14 do something that can't be undone, 85 change things and 9 only read.

Tools: 5 delete, 14 irreversible, 85 write, 9 read.
  • Reads private data · 1
  • Reads text others write · 76
  • Can send data out · 15
  • 16 findings

Tools listed without sign-in. Remote posture

With Anlyon

5 tools here can delete or cancel things and 14 do something that can't be undone. Govern them.

  • Anlyon lists the tools and adopts none.
  • Each tool you adopt gets approvals and receipts.
  • Your agent never holds the key.
Govern this server →

A tool your agent calls directly is not routed through Anlyon and is not governed by it. Anlyon Radar only lists and labels, and these labels are the same either way.

What the DPX — Settlement & Intelligence for AI Agents MCP server can do

The DPX — Settlement & Intelligence for AI Agents MCP server is published by the GitHub account untitledfinancial. It has 113 tools. It can delete or cancel things (5 tools: intelligence.subscription.delete, settlement.execute, settlement.nl, computer_use.pay and 1 more), do things that cannot be undone, such as sending, paying or publishing (stability.corridor, stability.settlement_window, stability.stablecoin_route, agent.audit_export and 10 more) and create or change data (85 tools: protocol.manifest, settlement.quote, esg.score, esg.batch and 81 more). 9 tools only read.

1 tool reads private data (search_docs); 76 tools bring in text other people can write (protocol.manifest, settlement.quote, esg.score and 73 more); 15 tools can send data out of your system (esg.watch, intelligence.subscribe, stability.corridor and 12 more). On its own it has all three parts of a prompt-injection exfiltration path: private data, text others write, and a way out.

The fixed checks matched 16 findings (annotation-conflict and credentials-mention). Each one comes with the text it matched.

Its remote endpoint lists its tools without sign-in, and its tools changed 2 times since Anlyon Radar first saw them.

Last scanned Nov 21, 2026, version 2.6.0.

Tools

Grouped by what each tool does, most consequential first. Open one for its reasons and full description; descriptions are the publisher's own words.

113 tools

Deletes or cancels 5Claude Code: deny
  • computer_use.payCost: $0.01 per call (x402, USDC on Base).Reads text others write1 finding

    Why it is in this group

    • The name "computer_use.pay" says "pay": a word that usually means money moves, which cannot be taken back.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.

    Findings

    • Info

      The description mentions "credentials".

      Matched text

      description: "...yment step. Call this instead of typing credentials into a UI. Describe what you see on scr..."

      What this check looks for
    Description, as the server wrote it
    Cost: $0.01 per call (x402, USDC on Base). Complete a payment that Claude's computer use session has identified on screen — a checkout form, wire transfer UI, invoice approval, or vendor portal payment step. Call this instead of typing credentials into a UI. Describe what you see on screen, provide the amount and recipient, and DPX runs the full oracle gate → compliance screen → settlement flow. Returns a receipt. Use whenever computer use encounters a payment that would otherwise require human re-entry or approval.

    Annotations the server declares

    • title: Computer-Use Payment
    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
  • intelligence.subscription.deleteCost: free.No findings

    Why it is in this group

    • The name "intelligence.subscription.delete" says "delete": a word for removing or ending something.
    • The server marks it destructive (destructiveHint: true).
    Description, as the server wrote it
    Cost: free. Cancel an intelligence subscription by ID. Stops future webhook alerts for that subscription. The alert log is retained for audit purposes.

    Annotations the server declares

    • title: Delete Intelligence Subscription
    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: true
    • openWorldHint: false
  • ramp.settleCost: $0.01 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • The input carries an amount (input.amount) and a currency, so it moves money.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.01 per call (x402, USDC on Base). Execute a DPX stablecoin settlement funded by a Ramp Agent Card — combines card creation and settlement in one call. Ramp handles the fiat conversion leg; DPX settles USDC or EURC on Base mainnet in ~30 seconds. Returns pacs.002 confirmation + SFDR PAI indicators. No crypto wallet pre-funding required. Requires Ramp account connected via ramp.connect.

    Annotations the server declares

    • title: Ramp Settlement
    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
  • settlement.executeCost: $0.01 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • The name "settlement.execute" says "execute": a word for a change.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.01 per call (x402, USDC on Base). Execute a DPX cross-border settlement. The Settlement Agent checks oracle conditions, reasons with Claude AI, and executes on-chain (or returns sandbox result if sandbox=true). Returns settlement ID, status (executed/held/sandbox/failed), tx hash, net amount, fees, oracle conditions, and AI reasoning. Default: sandbox=true — set sandbox=false only for live execution.

    Annotations the server declares

    • title: Execute Settlement
    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
  • settlement.nlCost: $0.01 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it destructive (destructiveHint: true).

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.01 per call (x402, USDC on Base). Execute a payment from a plain-English instruction. DPX's AI synthesis layer parses the instruction, runs the full oracle gate → compliance screen → settlement flow autonomously, and returns a receipt. Use this when the agent has a natural-language payment task rather than structured parameters. Examples: 'Pay Acme GmbH $25,000 for invoice #42', 'Send $10k to 0x... for vendor services', 'Settle the outstanding balance with Nova Trade SA'. Sandbox mode by default.

    Annotations the server declares

    • title: Natural Language Settlement
    • readOnlyHint: false
    • destructiveHint: true
    • idempotentHint: false
    • openWorldHint: true
Can't be undone 14Claude Code: ask
  • agent.audit_exportCost: free.Reads text others writeCan send data out1 finding

    Why it is in this group

    • The input addresses people (input.to): what it sends cannot be taken back.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a destination (input.to): what the tool writes goes there.

    Findings

    • Warn

      The server marks "agent.audit_export" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The input addresses people (input.to): what it sends cannot be taken back.

      What this check looks for
    Description, as the server wrote it
    Cost: free. Export an agent's own settlement audit trail: every settlement decided under its X-Agent-Id, with the KYA tier, mandate decision, policy decision, sanctions screen and oracle state recorded at the time, whether the paying wallet was the agent's registered wallet, and a SHA-256 digest of the export. Readable only by the agent's registered wallet: supply walletSignature, walletTimestamp and walletNonce — an EIP-191 signature over "DPX-KYA-AUDIT\n<agentId>\n<timestamp_ms>\n<nonce>" (see GET https://compliance.untitledfinancial.com/agent/auth-spec). Supports json or csv.

    Annotations the server declares

    • title: Agent Audit Export
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • flow_checkCost: free.Reads text others writeCan send data out1 finding

    Why it is in this group

    • The input addresses people (input.to): what it sends cannot be taken back.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a destination (input.to): what the tool writes goes there.

    Findings

    • Warn

      The server marks "flow_check" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":false,"openWorldHint":true}; The input addresses people (input.to): what it sends cannot be taken back.

      What this check looks for
    Description, as the server wrote it
    Cost: free. Single pre-flight call before settling. Runs oracle check, compliance screen, and stablecoin routing in parallel and returns a unified go/no-go decision. Replaces the 3-step oracle → screen → route loop. Returns: decision (PROCEED/HOLD/BLOCKED), recommended token, estimated net received, oracle score, compliance verdict, and a ready-to-use settleBody.

    Annotations the server declares

    • title: Flow Compliance Check
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • fx.cost_certaintyCost: $0.001 per call (x402, USDC on Base).Can send data out1 finding

    Why it is in this group

    • The input addresses people (input.to): what it sends cannot be taken back.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • outboundThe input takes a destination (input.to): what the tool writes goes there.

    Findings

    • Warn

      The server marks "fx.cost_certainty" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":false,"openWorldHint":false}; The input addresses people (input.to): what it sends cannot be taken back.

      What this check looks for
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). All-in settlement cost quote for cross-border payments. CFO-grade output: exact amount received in target currency after rail fees, 48h FX cost variance expressed in dollars, corridor stability overlay, and optimal execution window. Answers "if I send $X today, what does my counterparty receive net of everything, and how certain is that number?" Requires x402 micropayment.

    Annotations the server declares

    • title: FX Cost Certainty
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • fx.rateCost: free.Can send data out1 finding

    Why it is in this group

    • The input addresses people (input.to): what it sends cannot be taken back.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • outboundThe input takes a destination (input.to): what the tool writes goes there.

    Findings

    • Warn

      The server marks "fx.rate" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The input addresses people (input.to): what it sends cannot be taken back.

      What this check looks for
    Description, as the server wrote it
    Cost: free. Live mid-market FX rate for any currency pair. Returns mid rate, bid/ask spread, daily volatility %, regulatory flags for the corridor, and data freshness. Sourced from central bank rates (open.er-api.com, updated hourly, no API key required). Free.

    Annotations the server declares

    • title: FX Rate
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • invoice.createCost: free.Can send data outNo findings

    Why it is in this group

    • The name "invoice.create" says "create": a word for a change.
    • The input carries an amount (input.amount) and a currency, so it moves money.

    What it can reach

    • outboundThe input takes a URL (input.callbackUrl): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Cost: free. Create an agent-to-agent invoice. Agent A calls this to request payment from Agent B. Returns an invoiceId and payUrl — Agent B calls invoice.pay with the invoiceId to settle. Invoice expires after ttlSeconds (default 24h).

    Annotations the server declares

    • title: Create Invoice
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • invoice.payCost: $0.01 per call (x402, USDC on Base).No findings

    Why it is in this group

    • The name "invoice.pay" says "pay": a word that usually means money moves, which cannot be taken back.
    Description, as the server wrote it
    Cost: $0.01 per call (x402, USDC on Base). Pay an agent-to-agent invoice by ID. Retrieves the invoice, runs settlement via POST /settle, and marks the invoice PAID on success. In sandbox mode returns a simulated receipt; in live mode returns execution parameters for on-chain completion.

    Annotations the server declares

    • title: Pay Invoice
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • policy.checkCost: free.Can send data out1 finding

    Why it is in this group

    • The input addresses people (input.recipient): what it sends cannot be taken back.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • outboundThe input takes a destination (input.recipient): what the tool writes goes there.

    Findings

    • Warn

      The server marks "policy.check" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The input addresses people (input.recipient): what it sends cannot be taken back.

      What this check looks for
    Description, as the server wrote it
    Cost: free. Check a proposed payment against the agent's active policy before executing it. Returns ALLOW, HOLD, or BLOCK with a reason. Run this before every settlement call. ALLOW = proceed. HOLD = route to human review queue. BLOCK = halt, do not proceed.

    Annotations the server declares

    • title: Check Policy
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • ramp.agent_cardCost: $0.005 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • The input carries an amount (input.amount) and a currency, so it moves money.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.005 per call (x402, USDC on Base). Create a scoped Ramp Agent Card — a single-use virtual card with a merchant and amount cap, expires after first authorization or 12 hours. Used to fund the fiat leg of a DPX settlement without pre-funding a crypto wallet. Returns a task ID; poll ramp.agent_card_status to get PAN/CVV once ready. Requires cards:read_agentic scope (granted via ramp.connect).

    Annotations the server declares

    • title: Ramp Agent Card
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • ramp.compliance_screenCost: $0.002 per call (x402, USDC on Base).Reads text others write1 finding

    Why it is in this group

    • The input carries an amount (input.amount) and a currency, so it moves money.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.

    Findings

    • Warn

      The server marks "ramp.compliance_screen" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The input carries an amount (input.amount) and a currency, so it moves money.

      What this check looks for
    Description, as the server wrote it
    Cost: $0.002 per call (x402, USDC on Base). Compliance pre-screen for Ramp accounting agent payments — run before issuing an Agent Card to eliminate unnecessary human approval queues. Performs 5 checks in parallel: (1) FATF country risk on source and destination country, (2) amount threshold flags (CTR-equivalent at $10K, large-payment at $100K), (3) OpenSanctions global sanctions screen by counterparty name, (4) OpenSanctions PEP screen for individual counterparties or payroll, (5) GLEIF UBO chain with sanctions at each beneficial ownership node (if LEI provided). Returns APPROVED / FLAGGED / BLOCKED with a humanRequired boolean — true only for FLAGGED cases. APPROVED: issue card automatically, no human needed. BLOCKED: halt, do not proceed, do not notify counterparty. FLAGGED: route to compliance queue. Removes human-in-the-loop for the ~95% of payments that are clean.

    Annotations the server declares

    • title: Ramp Compliance Screen
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • receipt.createCost: free.Can send data outNo findings

    Why it is in this group

    • The name "receipt.create" says "create": a word for a change.
    • The input addresses people (input.recipient): what it sends cannot be taken back.

    What it can reach

    • outboundThe input takes a destination (input.recipient): what the tool writes goes there.
    Description, as the server wrote it
    Cost: free. Record a tamper-evident signed receipt for an agent financial action. Call immediately after every successful settlement. Returns a receipt ID and HMAC-SHA256 signature over the canonical receipt JSON — cryptographic proof the record has not been altered. Receipts are queryable by session or agent for audit.

    Annotations the server declares

    • title: Create Receipt
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • routeCost: $0.001 per call (x402, USDC on Base).Reads text others writeCan send data out1 finding

    Why it is in this group

    • The input addresses people (input.to): what it sends cannot be taken back.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a destination (input.to): what the tool writes goes there.

    Findings

    • Warn

      The server marks "route" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The input addresses people (input.to): what it sends cannot be taken back.

      What this check looks for
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Multi-stablecoin settlement routing. Given amount, source currency (from), and destination currency (to), returns all three stablecoin options (USDC, EURC, USDT) ranked by settlement efficiency. EURC is recommended for EUR destinations — eliminates cross-currency conversion. Returns settleBody ready to POST to /settle for the top-ranked option.

    Annotations the server declares

    • title: Route Payment
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • stability.corridorCost: $0.002 per call (x402, USDC on Base).Reads text others writeCan send data out1 finding

    Why it is in this group

    • The input addresses people (input.to): what it sends cannot be taken back.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a destination (input.to): what the tool writes goes there.

    Findings

    • Warn

      The server marks "stability.corridor" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The input addresses people (input.to): what it sends cannot be taken back.

      What this check looks for
    Description, as the server wrote it
    Cost: $0.002 per call (x402, USDC on Base). Corridor-specific settlement stability score (0–100) for any currency pair. Combines the live global Stability Oracle score with corridor-specific risk adjustments covering 28 currency pairs: regulatory flags (BCB/IOF for BRL, PBoC capital rules for CNH, BCRA controls for ARS, etc.), FX liquidity score based on active trading sessions at current UTC time, cascade penalty from live macro signals, and weekend/off-hours penalty. Returns SETTLE_NOW / DELAY_24H / DELAY_48H recommendation with rationale. Distinct from oracle.stability (which is global) and market.fx (which is spot-rate focused) — this answers "is this specific corridor safe to settle through right now?"

    Annotations the server declares

    • title: Stability Corridor
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • stability.settlement_windowCost: $0.002 per call (x402, USDC on Base).Reads text others writeCan send data out1 finding

    Why it is in this group

    • The input carries an amount (input.amount) and a currency, so it moves money.
    • The input addresses people (input.to): what it sends cannot be taken back.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a destination (input.to): what the tool writes goes there.

    Findings

    • Warn

      The server marks "stability.settlement_window" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The input carries an amount (input.amount) and a currency, so it moves money.

      What this check looks for
    Description, as the server wrote it
    Cost: $0.002 per call (x402, USDC on Base). Optimal settlement execution window analysis for a specific cross-border payment over the next 72 hours. Generates 18 × 4-hour time slots and scores each by composite risk: corridor stability, FX session liquidity, cascade level decay/growth based on macro outlook, weekend/off-hours penalty, and counterparty ESG tier (if LEI provided). Returns a ranked window schedule with OPTIMAL / GOOD / ACCEPTABLE / AVOID classification per slot, a best-window recommendation, and large-amount splitting guidance for settlements ≥ $5M. Use this before scheduling large cross-border settlements to minimize execution risk.

    Annotations the server declares

    • title: Settlement Window
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • stability.stablecoin_routeCost: free.Can send data out1 finding

    Why it is in this group

    • The input addresses people (input.to): what it sends cannot be taken back.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • outboundThe input takes a destination (input.to): what the tool writes goes there.

    Findings

    • Warn

      The server marks "stability.stablecoin_route" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The input addresses people (input.to): what it sends cannot be taken back.

      What this check looks for
    Description, as the server wrote it
    Cost: free. Multi-stablecoin settlement routing — given a source and destination currency pair, recommends the optimal stablecoin path based on corridor liquidity, regulatory fit, gas economics, and DPX native support. Returns a ranked list of stablecoins (USDC, EURC, BRLA, MXNC, NGNC, AEDX, PYUSD, USDT, and others) with regulatory flags, MiCA/GENIUS Act compliance status, liquidity tier, and warnings. Identifies blocked routes (e.g. USDT for EU under MiCA, BRLA before BCB Resolution 561 deadline). Use before settlement to avoid regulatory penalties and ensure optimal execution path.

    Annotations the server declares

    • title: Stablecoin Route
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
Changes things 85Claude Code: ask
  • agent.authorityCost: free.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: free. What is this DPX agent allowed to do right now? One read-only answer combining its KYA tier, its effective (tier-clamped) spend mandate, the spend policies held by the Policy Engine under its agentId or registered wallet, and the combined effective limits (the lowest cap wins). Also states what is enforced at settlement (the mandate, live) and what is advisory (policies, unless the deployment enables policy checks), and where the agent's wallet-signed audit trail lives.

    Annotations the server declares

    • title: Agent Authority
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • agent.kya_registerCost: free.Reads text others write1 finding

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.

    Findings

    • Info

      The input schema at properties.publicKey.description mentions "credential".

      Matched text

      inputSchema.properties.publicKey.description: "Optional public key for credential signature verification."

      What this check looks for
    Description, as the server wrote it
    Cost: free. KYA — Know Your Agent. Three-tier registration model — compliance burden scales with settlement risk, no documents ever required. ANONYMOUS: agent name only, $1K/day cap, instant. REGISTERED: add ownerEntity + ownerEmail (self-attested, no verification), $25K/day cap, instant. ownerLei (active GLEIF LEI) is recorded as a claim at REGISTERED tier — a LEI is public, so it does not prove ownership. VERIFIED requires an attestation from a trusted issuer naming this agentId and LEI (POST /agent/:id/kyc-attestation). VERIFIED agents get institutional caps (governed by mandate), FATF R.16 attestation on every settlement, and full AP2 mandate support. To bind a settlement wallet (needed to later authorize mandates, revocation and attestations), supply walletAddress with walletSignature, walletTimestamp and walletNonce — message formats at GET https://compliance.untitledfinancial.com/agent/auth-spec. Legal basis: FATF R.16 originator = owner entity (not the agent); MiCA Art. 45/72 accepts LEI; GENIUS Act satisfied by entity attestation.

    Annotations the server declares

    • title: Register Agent (KYA)
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • agent.kya_verifyCost: free.1 finding

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    Findings

    • Info

      The description mentions "credential".

      Matched text

      description: "...d DPX agent and receive a signed 1-hour credential. Returns KYA level, effective spend cap..."

      What this check looks for
    Description, as the server wrote it
    Cost: free. Verify a registered DPX agent and receive a signed 1-hour credential. Returns KYA level, effective spend caps (tier or mandate), owner verification status, mandate active status, and FATF R.16 compliance attestation. Attach credential.signature as X-Agent-Credential header and agentId as X-Agent-Id header on DPX /settle requests — enables mandate enforcement, per-agent audit trail, and FATF attestation. Credential expires in 1 hour; call again to refresh before expiry.

    Annotations the server declares

    • title: Verify Agent (KYA)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • agent.mandate_createCost: free.No findings

    Why it is in this group

    • The name "agent.mandate_create" says "create": a word for a change.
    Description, as the server wrote it
    Cost: free. Create or update an AP2-compatible spend mandate for a REGISTERED or VERIFIED DPX agent. Sets per-agent settlement constraints: max notional per settlement, daily cap, optional counterparty whitelist (LEIs or wallets), allowed currency pairs, ESG floor, and expiry. ANONYMOUS agents cannot hold mandates — register with ownerEntity + ownerEmail first. REGISTERED agents have mandate caps clamped to their tier limit ($25K). VERIFIED agents (GLEIF LEI confirmed) set their own caps with no platform ceiling. Mandate is AP2-formatted for interoperability with Google Agent Payments Protocol.

    Annotations the server declares

    • title: Create Spend Mandate
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • analytics.overviewCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Get live DPX performance analytics. Returns current stability score, ESG composite scores, live fee breakdown, oracle health across all data sources, and a settlement readiness assessment. Use for dashboards, reporting, and AI-driven monitoring of protocol health.

    Annotations the server declares

    • title: Analytics Overview
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • batch_settleCost: $0.01 per call (x402, USDC on Base).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Cost: $0.01 per call (x402, USDC on Base). Submit multiple settlements in a single call. Runs all settlements concurrently — one failure does not block others. Returns a summary (total/succeeded/failed) and per-item results mirroring what POST /settle would return. Maximum 50 per batch.

    Annotations the server declares

    • title: Batch Settlement
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • card.positionsCost: $0.001 per call (x402, USDC on Base).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Plan treasury settlement for a crypto card program. Accepts net positions per corridor (e.g. USD-BRL: $2.3M, USD-EUR: €450K) and returns an optimal settlement plan — which corridors to settle now vs. hold, which stablecoin to use per corridor, and estimated all-in fee. No settlement is executed. Call this before card.settle to review the plan. Free.

    Annotations the server declares

    • title: Card Positions
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • card.settleCost: $0.01 per call (x402, USDC on Base).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Cost: $0.01 per call (x402, USDC on Base). Execute treasury settlement for a crypto card program. Takes the same positions array as card.positions but executes all settlements via DPX batch — compliance-gated, oracle-priced, stablecoin-routed. Each position requires a recipientAddress. Use sandbox:true for testing. Returns per-corridor settlement results and a summary.

    Annotations the server declares

    • title: Card Settlement
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • compliance.aml_screenCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). AML Screen with Behavioural Profiling — deep AML assessment combining sanctions matching (OFAC, EU, UN, OFSI), behavioural pattern analysis, transaction velocity signals, and counterparty network risk. Returns CLEAR / REVIEW / BLOCK decision with confidence score, matched sanctions lists, behavioural anomaly flags, and a structured compliance narrative for audit. Use for high-value payments, onboarding, or any counterparty flagged by flow_check. Accepts wallet address or entity identifier. x402 paid, USDC on Base.

    Annotations the server declares

    • title: AML Screen
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • compliance.pep_screenCost: free.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: free. Screen an individual by name against the OpenSanctions PEP (Politically Exposed Person) dataset. PEPs include heads of state, senior government officials, senior executives of state-owned enterprises, senior politicians, senior military officers, judicial officials, and their close associates and family members. Returns match confidence, position/role, nationality, related entities, and an overall risk level (HIGH / MEDIUM / LOW / NONE). HIGH or MEDIUM matches require Enhanced Due Diligence (EDD) per FATF Recommendations 12 and 13 before settlement. Optionally filter by country.

    Annotations the server declares

    • title: PEP Screening
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • compliance.regulatory_calendarCost: free.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Cost: free. Returns a structured calendar of upcoming and in-effect compliance obligations across MiCA (EU crypto-asset markets regulation), SFDR (Sustainable Finance Disclosure Regulation), CSRD (Corporate Sustainability Reporting Directive), the US GENIUS Act (payment stablecoin framework), and FATF Recommendations 15/16. For each event: framework, jurisdiction, requirement summary, effective date, impact level, and article reference. Also returns a DPX alignment section mapping each framework to the specific DPX endpoints that satisfy it. Use this before settlement workflow design, compliance gap analysis, or regulatory reporting.

    Annotations the server declares

    • title: Regulatory Calendar
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • compliance.sfdr_pai_reportCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). SFDR Annex I PAI Report — generates the full Principal Adverse Impact statement for a portfolio of up to 50 entities (by LEI or wallet address). Returns all 18 mandatory indicators plus 2 optional climate indicators in the SFDR Annex I table format. Required for SFDR Article 8 and Article 9 product annual disclosure. Accepts LEIs or company names (GLEIF-resolved). x402 paid, USDC on Base.

    Annotations the server declares

    • title: SFDR Annex I PAI Report
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • compliance.sfdr_screenCost: $0.01 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.01 per call (x402, USDC on Base). SFDR Article 8/9 Compliance Screen — determines whether a financial product or counterparty meets the Sustainable Finance Disclosure Regulation classification thresholds. Returns Article classification (6/8/9), Principal Adverse Impact indicator flags across 18 mandatory PAI metrics, ESG score alignment, and DNSH (Do No Significant Harm) assessment. Required for EU-regulated funds and any payment counterparty with SFDR disclosure obligations. Pass LEI or wallet address. x402 paid, USDC on Base.

    Annotations the server declares

    • title: SFDR Article 8/9 Screen
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • compliance.ubo_chainCost: free.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: free. Trace the beneficial ownership chain for any legal entity up to 3 levels deep using GLEIF relationship records, then screen every node in the chain against the OpenSanctions consolidated sanctions list (OFAC SDN, EU, UN, UK OFSI). Returns chain structure (SUBJECT → DIRECT_PARENT → ULTIMATE_PARENT), per-node sanctions status, LEI lapse flags, overall CLEAR / REVIEW_REQUIRED / BLOCKED verdict, and FATF R.16 beneficial ownership compliance attestation. Required for correspondent banking due diligence, FATF R.12/13 UBO identification, and MiCA Article 72 counterparty risk management.

    Annotations the server declares

    • title: UBO Chain Lookup
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • compliance.vendor_riskCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Composite Vendor Risk Score — combines compliance (65%) and ESG (35%) signals into a single vendor risk assessment. Returns an overall risk tier (LOW / MEDIUM / HIGH / CRITICAL), component breakdown, sanctions status, PEP flags, ESG pillar scores, and a payment recommendation (PROCEED / PROCEED_WITH_MONITORING / HOLD / BLOCK). Designed for pre-payment vendor due diligence, treasury onboarding, and supplier risk review. Accepts LEI or wallet address. x402 paid, USDC on Base.

    Annotations the server declares

    • title: Vendor Risk Score
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • compliance.vop_verifyCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Verification of Payee — cryptographically verified payee check before a cross-border payment. Resolves a wallet address against GLEIF LEI registry, ESG Oracle, and on-chain activity to confirm the payee is who they claim to be. Returns a verification attestation, match confidence score, identity fields confirmed, and a signed verification receipt. Required for FATF R16 Travel Rule compliance and recommended for any payment over $3,000. $0.075 USDC per verification.

    Annotations the server declares

    • title: Verification of Payee
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • compute.costCost: $0.001 per call (x402, USDC on Base).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Get a model recommendation for a task type without running inference. Returns the best free model for the task, its strengths and speed tier, and a list of alternatives. Use this when an agent needs to select a model before committing to inference, or to surface model selection logic to a human. Free.

    Annotations the server declares

    • title: Compute Cost
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • compute.modelsCost: $0.001 per call (x402, USDC on Base).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). List all AI models available through DPX Compute. All models are free-tier (no token cost) — routed via OpenRouter. Returns model IDs, provider, capability strengths, context window, and speed tier. Use this before compute.route to understand what models are available and pick the right one for a task. Free.

    Annotations the server declares

    • title: Compute Models
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • compute.routeCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Route a task to the best available free AI model and run inference. DPX selects the model based on the task type (reasoning → DeepSeek R1, code → Llama 3.3 70B, multilingual → Qwen 2.5 72B, fast → Llama 3.1 8B), calls OpenRouter, and returns the completion. All models are free-tier — no token cost. Pay per call in USDC via x402. Use this when an agent needs to delegate a subtask to a language model without managing model selection or API keys.

    Annotations the server declares

    • title: Compute Routing
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • defi.research_papersCost: $0.25 per call (x402, USDC on Base).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Cost: $0.25 per call (x402, USDC on Base). 13 structured institutional research papers on CBDCs, stablecoins, crypto regulation, and cross-border payments — from BIS, the Federal Reserve, ECB, FSB, IMF, G7, SWIFT, and R3 — each with a summary and key findings. These papers are free but otherwise scattered as individual PDFs with no query layer; this makes them queryable in one place. Optional keyword filter.

    Annotations the server declares

    • title: CBDC/Stablecoin Research Library
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • dpx.metricsCost: free.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Cost: free. Live performance metrics for the DPX settlement infrastructure — pulled directly from production telemetry. Returns request volumes, error rates, growth trends, per-service breakdown, and spike analysis across all active DPX workers. Free — designed for investor due diligence, analyst queries, and Standard Metrics / portfolio management integrations. No auth required.

    Annotations the server declares

    • title: DPX Metrics
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • esg.batchCost: $0.002 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.002 per call (x402, USDC on Base). Screen up to 50 entities in a single call. Accepts LEIs or company names (GLEIF-resolved). Returns results ranked by composite ESG score descending — highest scoring counterparties first. Useful for portfolio-level compliance screening, supplier due diligence, and TMS pre-payment checks. Name resolution is slower than direct LEI input.

    Annotations the server declares

    • title: ESG Batch Score
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • esg.portfolioCost: $0.005 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.005 per call (x402, USDC on Base). Score an entire counterparty portfolio in one call (up to 200 entities by LEI or name). Returns portfolio-level composite E/S/G scores, tier distribution, aggregate fee surcharge impact in basis points, worst offenders (bottom 10% by composite), top performers, MiCA Article 72 ongoing monitoring status, and SFDR PAI flags. The canonical pre-settlement compliance check for treasury systems and TMS integrations.

    Annotations the server declares

    • title: ESG Portfolio Score
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • esg.scoreCost: $0.10 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.10 per call (x402, USDC on Base). Get the live counterparty risk score (ESG-denominated) for a wallet address or the protocol default. Returns Environmental, Social, and Governance risk scores (0–100 each), composite weighted average, and the compliance-adjusted settlement fee percentage this score produces. Updated hourly from 6 institutional data sources: WorldBank, IMF, OECD, UN SDG API, ClimateMonitor, and SEC EDGAR. Required by EU SFDR Principal Adverse Impact reporting and CSRD financed emissions disclosure for institutional clients.

    Annotations the server declares

    • title: ESG Score
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • esg.supply_chainCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Supply Chain ESG Screen — screens an entire supplier/counterparty supply chain for ESG risk exposure. Accepts a focal entity (by LEI or wallet address) and traces upstream supply chain relationships via GLEIF. Returns per-tier ESG scores, highest-risk nodes, SFDR PAI exposure from supply chain, and a Scope 3 financed-emissions estimate. Useful for procurement teams, ESG-adjusted settlement pricing, and CSRD supply chain due diligence. x402 paid, USDC on Base.

    Annotations the server declares

    • title: Supply Chain ESG Screen
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • esg.trendCost: free.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Cost: free. Get the historical ESG composite trend for a specific entity by LEI. Returns score history, trend direction (IMPROVING / STABLE / DETERIORATING), and delta over the requested window. Data accumulates each time the entity is scored via esg.lookup, esg.batch, or esg.portfolio. Useful for due diligence, MiCA Article 72 ongoing monitoring reports, and detecting counterparties whose ESG posture is degrading.

    Annotations the server declares

    • title: ESG Trend
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • esg.watchCost: free.Can send data outNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.

    What it can reach

    • outboundThe input takes a URL (input.webhookUrl): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Cost: free. Register an entity for ongoing ESG monitoring. DPX checks the score daily and fires a webhook when the composite score shifts by ≥ thresholdPoints. Satisfies MiCA Article 72 ongoing monitoring requirements. Returns a watchId for status checks and cancellation. Webhook payload includes previous/current score, delta, and tier change.

    Annotations the server declares

    • title: ESG Watch
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • fees.scheduleCost: free.Reads text others write1 finding

    Why it is in this group

    • The name "fees.schedule" says "schedule": a word for a change.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.

    Findings

    • Warn

      The server marks "fees.schedule" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The name "fees.schedule" says "schedule": a word for a change.

      What this check looks for
    Description, as the server wrote it
    Cost: free. Get the complete DPX fee schedule: all components (core/FX/ESG/license) read from the on-chain router, how the ESG fee follows the paying agent's own score, and scenario examples.

    Annotations the server declares

    • title: Fee Schedule
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • fees.verifyCost: $0.001 per call (x402, USDC on Base) (scales with payment amount above $10K, max $5.00).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base) (scales with payment amount above $10K, max $5.00). Verify that the off-chain fee quote matches what the on-chain DPXSettlementRouter contract will charge. Returns feesMatch (true/false). Call after get_quote and before settle to confirm fee integrity.

    Annotations the server declares

    • title: Verify Fees
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • forecast.calendarCost: free.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Cost: free. Seasonal climate event calendar with commodity impact. Returns 12 critical annual windows (hurricane season, corn pollination, Brazil frost risk, Black Sea harvest, ENSO influence periods, etc.) sorted by urgency — active windows first, then by months until next occurrence. Each entry includes affected commodities, severity, and the agronomic basis.

    Annotations the server declares

    • title: Forecast Calendar
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • forecast.commodity_outlookCost: $0.002 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.002 per call (x402, USDC on Base). Climate-driven price pressure outlook for a commodity. Returns BULLISH/BEARISH/NEUTRAL signal with 30/60/90-day horizons, confidence score, per-region stressor breakdown, and current FRED price reference. Covers 11 commodities: WHEAT, CORN, SOYB, COFFEE, COCOA, COTTON, SUGAR, WTI, NG, COPPER, LUMBER. Designed for institutional research teams evaluating commodity positions. Signals reflect supply constraint risk from climate — not a financial recommendation. Cache: 4h.

    Annotations the server declares

    • title: Commodity Outlook Forecast
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • forecast.heat_checkCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Commodity Heat Check — all 11 commodity symbols simultaneously scored RED / YELLOW / GREEN. Fast single-call scan of the full commodity universe: WHEAT, CORN, SOYB, COFFEE, COCOA, COTTON, SUGAR, WTI, NG, COPPER, LUMBER. Returns per-symbol climate stress signal, risk driver, and urgency label. Ideal for daily monitoring, dashboard widgets, and as a triage step before calling forecast.commodity_outlook for detail on flagged symbols. Free.

    Annotations the server declares

    • title: Commodity Heat Check
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • forecast.portfolio_stressCost: $0.003 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.003 per call (x402, USDC on Base). Climate stress test for a multi-commodity portfolio. Pass up to 20 positions with weights (percentages or fractions — normalized internally). Returns aggregate portfolio climate score, which positions are most stressed, which could act as climate hedges, and a plain-language summary. Useful for commodity fund managers evaluating aggregate climate exposure before rebalancing.

    Annotations the server declares

    • title: Portfolio Stress Forecast
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • forecast.production_regionsCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). All ~40 global commodity production regions ranked by current climate risk score. Each region shows which commodities it affects and its current climate risk level (HIGH/MODERATE/LOW). Use to identify which geographic zones are under active climate stress and which commodities are most exposed.

    Annotations the server declares

    • title: Production Regions Forecast
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • forecast.scenarioCost: $0.002 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.002 per call (x402, USDC on Base). What-if climate scenario analysis. Apply a named scenario or custom stressor multipliers to any subset of commodities and see how signals shift. Built-in scenarios: la_nina_moderate, la_nina_severe, el_nino_moderate, gulf_hurricane_major, us_plains_drought_severe, black_sea_disruption, brazil_frost, chile_drought_copper, pacific_northwest_wildfire. Use to stress-test a commodity thesis before committing to a position.

    Annotations the server declares

    • title: Scenario Forecast
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • forecast.subscribeCost: $0.001 per call (x402, USDC on Base).Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "forecast.subscribe" says "subscribe": a word for a change.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    • outboundThe input takes a URL (input.webhookUrl): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Commodity Alert Webhook — register a URL to receive push alerts when a commodity signal crosses a threshold. Supports alerts on signal changes (BULLISH → BEARISH), confidence drops below a floor, or climate stress score crossing a level. Returns a subscriptionId for management. Alerts fire within 30 minutes of a signal change. Useful for autonomous agents that need to react to commodity market shifts without polling. x402 paid (first month), USDC on Base.

    Annotations the server declares

    • title: Commodity Alert Subscription
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • forecast.tcfd_reportCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). TCFD Physical Risk Report — generates a Task Force on Climate-related Financial Disclosures (TCFD) physical risk assessment for a commodity portfolio. Maps climate signals from all 11 commodity markets to TCFD physical risk categories (acute and chronic), scores transition risk alignment, and produces a structured report suitable for investor disclosure. Pass portfolio as an array of {symbol, positionUsd}. x402 paid, USDC on Base.

    Annotations the server declares

    • title: TCFD Physical Risk Report
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • fx.corridorsCost: free.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Cost: free. All 60+ currency corridors with current stability tier, daily volatility estimate, and regulatory flags. Sort is best-first (OPTIMAL → ADVERSE). Use to compare corridors before choosing a payment route — e.g. "which LATAM corridor is most stable for a $2M payment this week?" Filter by source currency with the from parameter. Free.

    Annotations the server declares

    • title: FX Corridors
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • integration.kyribaCost: $0.001 per call (x402, USDC on Base).1 finding

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.

    Findings

    • Info

      The description mentions "credentials".

      Matched text

      description: "...reakdown. The client's own Kyriba OAuth credentials are used — DPX never holds credentials...."

      What this check looks for
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Kyriba TMS Integration — DPX settlement bridge for Kyriba treasury management systems. Accepts a Kyriba-formatted payment instruction and executes it through the DPX settlement flow with oracle gating, compliance screening, and stablecoin routing. Returns a Kyriba-compatible confirmation response including settlement ID, execution status, and fee breakdown. The client's own Kyriba OAuth credentials are used — DPX never holds credentials. Useful for treasury teams running Kyriba who want DPX oracle intelligence and on-chain settlement without leaving their TMS workflow. Free (settlement x402 fee charged at execute step).

    Annotations the server declares

    • title: Kyriba TMS Integration
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • integration.travel_ruleCost: $0.001 per call (x402, USDC on Base).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). FinCEN Travel Rule Compliance — generates and transmits Travel Rule records for cross-border payments over $3,000. Creates a FATF R16-compliant record with originator and beneficiary identity data, transaction details, and a cryptographic attestation. Returns a Travel Rule ID for the audit trail. Accepts the same settlement parameters as settlement.execute — use after oracle and compliance gates pass. Can also receive inbound Travel Rule records from counterparty VASPs. Free (included in settlement flow; standalone record creation available).

    Annotations the server declares

    • title: Travel Rule Compliance
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • integration.verifyCost: free.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Cost: free. Initiate a $0.01 USDC onboarding handshake for a new DPX integration. Returns the DPX treasury address and payment instructions. The client sends $0.01 USDC on Base mainnet to confirm their wallet is funded and settlement rails are clear. Call integration.status to poll for confirmation. Required for all new integrations before production settlements are enabled.

    Annotations the server declares

    • title: Verify Integration
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • intelligence.48h_callCost: $0.001 per call (x402, USDC on Base).Reads text others write1 finding

    Why it is in this group

    • The name "intelligence.48h_call" says "call": a word for a change.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.

    Findings

    • Warn

      The server marks "intelligence.48h_call" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

      Matched text

      annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The name "intelligence.48h_call" says "call": a word for a change.

      What this check looks for
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). 48-Hour Forward Macro Call — structured 48h forward view with the top 3 macro risks for the next 48 hours, per-corridor settlement guidance, and a recommended execution strategy for large cross-border payments. Distinct from composite (which covers a 30/60/90d horizon) — this is the short-horizon tactical call for agents with imminent settlement decisions. Returns: topRisks (ranked by probability × impact), corridorAdvice (SETTLE_NOW / DELAY / AVOID per major pair), executionWindow (best 4h window in next 48h), settlementAdvice (structured go/no-go), and AI synthesis. $0.50 USDC. GET, no input required.

    Annotations the server declares

    • title: 48-Hour Forward Macro Call
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • intelligence.aftershockCost: $0.75 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.75 per call (x402, USDC on Base). Aftershock Intelligence — models the secondary waves that follow a primary cascade event. Takes a primary shock (origin node, event type, magnitude, elapsed hours) and returns three aftershock waves: Wave 1 (0–72h immediate secondary effects), Wave 2 (1–4 weeks policy response distortions), Wave 3 (1–6 months structural changes now permanently locked in). Identifies which nodes are rebounding, which face amplified pressure, and which are structurally altered. Companion to market.cascade — run cascade first, then aftershock to see the full picture. POST with origin, eventType, magnitude, elapsedHours.

    Annotations the server declares

    • title: Aftershock Intelligence
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • intelligence.butterflyCost: $0.75 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.75 per call (x402, USDC on Base). Butterfly Effect Cascade — POST variant of market.cascade. Models how a shock originating at a specific node propagates through a 25-node network. Distinct from market.cascade (which takes a free-text event and models broad macro impact) — butterfly requires a specific origin node from the node taxonomy and returns a directed, time-ordered cascade chain optimised for agent decision loops. Use when you have a precise origin event and need to trace which downstream nodes are hit in what order. Input: originNode (required), magnitude 1–10, elapsed hours. $0.75 USDC. POST.

    Annotations the server declares

    • title: Butterfly Effect Cascade (POST)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • intelligence.climateCost: $1.25 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $1.25 per call (x402, USDC on Base). Climate Intelligence Suite — three climate endpoints bundled. (1) climate: 50-year precipitation shift across 10 agricultural zones — food inflation and migration pressure signals. (2) climate-pulse: near-real-time temperature anomaly and drought stress across 8 US regions, updated every 3h — use for commodity trading desks and ag-sector settlement exposure. (3) earth-systems: Earth Health Index (0–100) and proximity to 9 climate tipping points — long-horizon investors and TNFD-aligned portfolio stress. All three returned in one response. GET, no input required. $0.50 USDC (bundled rate). Data: NOAA, USDA, FRED. Cache: climate 6h, pulse 3h, earth-systems 24h.

    Annotations the server declares

    • title: Climate Intelligence Suite
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • intelligence.compositeCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Composite Intelligence Report — synthesizes all active intelligence signals into a single structured macro brief. Calls all non-POST intelligence endpoints internally, weights their outputs by a live calibration model, and returns a ranked risk register, corridor-level settlement guidance, a 30/60/90-day outlook, and a full AI synthesis narrative. The highest-value single-call output in the intelligence API — use when a counterparty, treasury team, or agent needs a comprehensive cross-domain briefing rather than individual signals. Also available: synthesis (same output structured as a formatted executive report rather than machine-readable JSON). $1.00 USDC. GET, no input required.

    Annotations the server declares

    • title: Composite Intelligence Report
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • intelligence.contagionCost: $0.75 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.75 per call (x402, USDC on Base). Contagion Intelligence — simulates how a macro or financial shock spreads through 30 nodes across 6 domains (financial systems, real economies, commodity networks, policy anchors, social systems, physical infrastructure) using an epidemiological R-value model. Returns system R trajectory, per-epoch spread map, superspreader nodes, containment forecast, and AI briefing. R < 1.0 = self-limiting; R ≥ 1.0 = expanding. Call /contagion/nodes first to discover valid origin IDs. POST with origin and magnitude.

    Annotations the server declares

    • title: Contagion Network Intelligence
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • intelligence.esg_reportingCost: $1.75 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $1.75 per call (x402, USDC on Base). ESG Regulatory Reporting Suite — four SFDR/TNFD/EU Taxonomy endpoints bundled into one call. Returns everything needed for institutional ESG disclosure in a single response. (1) sfdr-dashboard: SFDR PAI indicators for a portfolio address — mandatory Principal Adverse Impact metrics under SFDR Article 7. (2) financed-emissions: PCAF-aligned financed emissions calculation — Scope 3 Category 15, required under CSRD and SFDR PAI 1. (3) taxonomy-alignment: EU Taxonomy alignment scoring — % of activities meeting Do No Significant Harm criteria across 6 environmental objectives. (4) tnfd-report: full TNFD LEAP report — nature-related risk and opportunity assessment across Locate, Evaluate, Assess, Prepare stages. $1.50 USDC (bundled, saves $1.00 vs individual calls). POST for portfolio address.

    Annotations the server declares

    • title: ESG Regulatory Reporting (SFDR/TNFD/EU Taxonomy)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • intelligence.gender_riskCost: $0.75 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.75 per call (x402, USDC on Base). Gender Risk & Opportunity Intelligence — maps the structural relationship between GBV prevalence, legal discrimination, female labour force participation, and economic outcomes across 18 countries. Returns two independent scores: gbvRiskScore (0–100 suppression risk — high GBV → female LFPR suppression → GDP drag → fiscal stress → sovereign risk premium) and opportunityScore (0–100 reform upside — improving GBV indicators, closing LFPR gender gaps, and strengthening legal rights precede FDI inflows and consumer credit expansion). Five transmission mechanisms. Live FRED economic stress feedback. AI synthesis. Data: WHO GHO, World Bank WDI, FRED. 12h cache. No input required — GET.

    Annotations the server declares

    • title: Gender Risk Intelligence
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • intelligence.macro_stressCost: $0.90 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.90 per call (x402, USDC on Base). Macro Stress & Instability Intelligence — two endpoints bundled into one tool call. (1) Macro-stress: credit regime classification (EXPANSION / LATE_CYCLE / STRESS / CRISIS) derived from FRED yield curve, credit spreads, and money supply — provides a 2–6 week FX lead signal. (2) Instability: structural regime-level risk across US, China, Russia, and EU — not episodic events but foundational political-economic stability. Use this before large cross-border settlements or when macro conditions are a factor in a payment decision. GET, no input required. $0.25 USDC (bundled rate, covers both signals). Data: FRED. 1h cache.

    Annotations the server declares

    • title: Macro Stress & Instability
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • intelligence.nature_riskCost: $0.50 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.50 per call (x402, USDC on Base). Nature & Water Risk Intelligence — two TNFD-aligned endpoints bundled. (1) water-risk: physical water stress across 6 regions with sector-level exposure for agriculture, semiconductors, and beverages — TNFD LEAP aligned, relevant to ESG screening and climate-adjusted settlement. (2) biodiversity: TNFD-aligned nature risk across 8 regions — IUCN Red List signals, sector dependency/impact matrix, SFDR PAI 7 biodiversity flag. Together they cover the nature-related financial disclosure requirements under TNFD and SFDR. GET, no input required. $0.35 USDC (bundled rate). Data: USGS, FAO, IUCN, World Bank. 12h cache.

    Annotations the server declares

    • title: Nature & Water Risk (TNFD)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • intelligence.political_riskCost: $1.25 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $1.25 per call (x402, USDC on Base). Political & Climate Transition Risk — two forward-looking structural risk endpoints bundled. (1) political-risk: regime stability, policy risk, and sanctions exposure across major economies — structured assessment of political disruption to cross-border payment corridors. (2) transition-risk: climate policy transition risk across industries and geographies — carbon price exposure, stranded asset risk, policy velocity, regulatory tightening signals. Critical for SFDR Article 8/9 compliance and climate-adjusted settlement pricing. GET, no input required. $0.75 USDC (bundled rate). Data: ACLED, FRED, World Bank WGI. 6h cache.

    Annotations the server declares

    • title: Political & Transition Risk
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • intelligence.resonanceCost: $0.75 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.75 per call (x402, USDC on Base). Resonance Intelligence — detects when multiple independent macro forces are oscillating in phase across 28 signals in 5 domains, amplifying each other rather than cancelling. A single shock is manageable; resonance turns a bad quarter into a systemic crisis. Returns per-signal phase angles, resonance clusters (groups of 3+ aligned signals), amplitude amplification factor, system resonance score (0–100), and historical danger-zone comparison to crisis precedents (2008, 2011, 2020, 1997 EM). No input required — GET.

    Annotations the server declares

    • title: Resonance Phase Intelligence
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • intelligence.sovereign_debtCost: $0.50 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.50 per call (x402, USDC on Base). Sovereign Debt & Currency Stress Intelligence — two FX-adjacent macro endpoints bundled. (1) sovereign-debt: debt sustainability and default risk for US, EU, UK, and Japan — tracks debt/GDP trajectory, yield dynamics, and cross-border contagion risk relevant to corridor safety. (2) currency-stress: G10 + EM currency stress scores, dollar regime classification, capital flow direction, and per-corridor settlement impact — which currency pairs are under structural pressure beyond current FX rates. Complements market.fx (which covers spot execution risk) with underlying structural analysis. GET, no input required. $0.35 USDC (bundled rate). Data: FRED, IMF. 2h cache.

    Annotations the server declares

    • title: Sovereign Debt & Currency Stress
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • intelligence.subscribeCost: free.Can send data outNo findings

    Why it is in this group

    • The name "intelligence.subscribe" says "subscribe": a word for a change.

    What it can reach

    • outboundThe input takes a URL (input.webhookUrl): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Cost: free. Register a webhook to receive alerts when a DPX intelligence signal crosses a threshold. Supported signals: stability (overall 0–100 score), cascade (shock propagation risk), macro_stress, climate, fx, or any. The cron checks hourly and fires the webhook on crossing — edge-triggered, not repeated every hour. Returns a subscriptionId for status checks and cancellation. Use for treasury alert systems, TMS integrations, or autonomous agent monitoring loops.

    Annotations the server declares

    • title: Subscribe to Intelligence Alerts
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • intelligence.supply_chainCost: $0.50 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.50 per call (x402, USDC on Base). Supply Chain & Energy Transition Intelligence — two structural economic endpoints bundled. (1) supply-chain: global supply chain pressure index with waterway chokepoint scoring across 6 shipping lanes (Suez, Panama, Strait of Hormuz, Malacca, Bosphorus, Cape of Good Hope) — critical for trade-finance and goods-backed payment corridors. (2) energy-transition: renewable generation share, fossil fuel demand curve, and grid carbon intensity for US, EU, and global — relevant for climate-risk-adjusted treasury allocation and SFDR PAI reporting. GET, no input required. $0.25 USDC (bundled rate). Data: FRED, EIA. 4h cache.

    Annotations the server declares

    • title: Supply Chain & Energy Transition
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • intelligence.tectonicCost: $0.75 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.75 per call (x402, USDC on Base). Tectonic Intelligence — maps slow-moving structural stress across 22 fault lines in 5 domains (demographic, fiscal, environmental, infrastructure, geopolitical). Each node carries current stress (0–100), accumulation rate (%/yr), tipping threshold, and estimated years to rupture. Where market.cascade traces an acute shock, tectonic surfaces latent pressure before it ruptures. Returns per-node stress state, rupture sequence, horizon timeline, and AI synthesis briefing. No input required — GET.

    Annotations the server declares

    • title: Tectonic Stress Intelligence
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • ledger.sessionCost: free.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Cost: free. Get the aggregated payment graph for a multi-agent session. Returns total USD moved, transaction count, and a chronological list of all payments made during the session. Use for cost accounting, audit, or to show a human what an agent run spent.

    Annotations the server declares

    • title: Ledger Session
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • market.cascadeCost: $0.002 per call (x402, USDC on Base) (scales with payment amount above $10K, max $5.00).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.002 per call (x402, USDC on Base) (scales with payment amount above $10K, max $5.00). Butterfly Effect Cascade Intelligence — models how a shock in one macro domain propagates through the interconnected web of climate, geopolitical, economic, and commodity systems. Given an origin event (e.g. armed conflict escalation, agricultural drought, central bank rate decision, rare earth export restriction) and a magnitude score, returns a time-ordered cascade chain showing which downstream systems are hit, in what sequence, with what attenuated signal strength, and an AI synthesis briefing on the highest-impact transmission paths. Covers 24 nodes across 4 domains: climate (drought, flood, carbon price, wildfire, sea-level stress, heatwave), geopolitical (sanctions, conflict, trade tariffs, regime change, election shock, port blockade), economic (rate decisions, inflation, sovereign debt, banking stress, currency crisis, recession), and commodity (oil, gas, grain, rare earth/lithium, copper, water, fertilizer). Purely macro intelligence — no settlement or stablecoin mechanics.

    Annotations the server declares

    • title: Market Cascade (Butterfly Effect)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • market.fxCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). FX Settlement Corridor Intelligence — per-pair execution risk assessment for 10 major currency corridors against USD: EUR, GBP, JPY, CAD, AUD, CHF, MXN, BRL, CNY, INR. Maps live FRED spot rates to settlement advice for each pair: SETTLE_NOW / SETTLE_WITH_HEDGE / DELAY_SHORT / DELAY_REVIEW / AVOID. Returns DXY dollar regime (STRONG_DOLLAR / NORMAL / WEAK_DOLLAR), regional block risk rollup (G4, Americas, Asia-Pacific), best corridors to settle through now, worst corridors to avoid or hedge, and recommended actions. Distinct from oracle.stability (which covers peg deviation and macro settlement gates) — this tool answers "which currency pairs are risky to settle through right now?" Data: FRED spot rates (DEXUSEU, DEXUSUK, DEXJPUS, etc.), DXY (DTWEXBGS). 1h cache.

    Annotations the server declares

    • title: FX Market Intelligence
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • market.shippingCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Shipping & Logistics Stress Intelligence — composite view of global freight market conditions across ocean, air, truck, and rail. Tracks energy-driven shipping costs (Brent crude, diesel), 8 key global trade routes with disruption status, and trade flow signals. Returns a settlementRelevance section mapping logistics conditions to cross-border payment corridor risk: invoice delay risk, trade finance stress, and affected corridors. Useful for treasury teams with supply chain financing exposure, trade finance desks, and agents pricing cross-border payments on goods-backed corridors. Data: FRED (Brent crude), EIA (US diesel). 4h cache.

    Annotations the server declares

    • title: Shipping & Logistics Stress
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • oracle.chaos_scoreCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Global Chaos Score — composite 0–100 instability index derived from all 11 Stability Oracle signal layers. Returns the raw numeric score, a regime label (CALM / ELEVATED / VOLATILE / HIGH / CRITICAL), a per-layer contribution breakdown, trend direction (IMPROVING / STABLE / DETERIORATING), and the key drivers pushing the score up or down. Distinct from oracle.stability (which gives a STABLE/CAUTION/UNSTABLE verdict) — this gives the underlying numeric signal useful for dashboards, monitoring, and threshold-triggered automation. Free.

    Annotations the server declares

    • title: Global Chaos Score
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • oracle.governanceCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Get the live governance score (0–100) for any legal entity identified by LEI or company name. Pulls from GLEIF (LEI registration status, renewal compliance) and World Bank Worldwide Governance Indicators (Government Effectiveness, Control of Corruption, Rule of Law). Returns composite governance score, tier (STRONG / ADEQUATE / MODERATE / WEAK / POOR), MiCA compliance flag, and per-source component breakdown. Complements esg.score by isolating the G pillar as a standalone institutional-grade signal.

    Annotations the server declares

    • title: Oracle Governance
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • oracle.myceliumCost: $0.75 per call (x402, USDC on Base) (scales with payment amount above $10K, max $5.00).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.75 per call (x402, USDC on Base) (scales with payment amount above $10K, max $5.00). Mycelium Network Oracle — models the global financial system as a living network and detects crisis formation from network topology before it surfaces in market data, typically 6–14 weeks ahead. Maps nodes (markets, economies, funding markets), threads (capital flow channels, correspondent banking, trade finance), nutrient flow (liquidity), stress signals (spread widening, FX stress), and dead zones (sanctioned corridors, failed correspondent networks). Returns network health score (0–100), regime classification (HEALTHY / THINNING / STRESSED_CONNECTIVITY / DEAD_ZONE_FORMING / FRUITING_BODY_IMMINENT), node-by-node connectivity, thread health, signal propagation speed, and fruiting body risk — the probability of a visible crisis with estimated lead time in weeks. Data: FRED (funding markets, credit spreads), BIS SDMX API (credit-to-GDP gaps), IMF DOTS (bilateral trade volumes). The only oracle that reads network topology rather than individual metrics.

    Annotations the server declares

    • title: Mycelium Network Oracle
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • oracle.railsCost: free.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: free. Get live health status of local payment rails relevant to a settlement. Returns per-rail status (OPERATIONAL/DEGRADED/DOWN), latency, last incident, and a composite health score. Key rails: PIX (Brazil), SEPA (Europe), FedACH (US domestic), CHAPS (UK), UPI (India), PromptPay (Thailand). Call this before domestic or regionally-specific settlements to confirm the destination rail is healthy.

    Annotations the server declares

    • title: Rail Status Oracle
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • oracle.settlement_gateCost: $0.10 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.10 per call (x402, USDC on Base). Settlement Gate — composite pre-flight check that bundles three clearance signals into one verdict. Calls: (1) Stability Oracle — macro/peg/FX gate; (2) Compliance Oracle — AML/sanctions tier for the counterparty; (3) SFDR pre-assessment — ESG PAI flags if applicable. Returns a single PROCEED / CAUTION / HOLD decision with per-gate sub-results and a plain-language rationale. Use this as a one-call replacement for running oracle.stability + flow_check + compliance.sfdr_screen separately before a large settlement. Pass counterparty wallet and settlement amount. Free.

    Annotations the server declares

    • title: Settlement Gate
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • oracle.stabilityCost: $0.10 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.10 per call (x402, USDC on Base). Get live macro stability assessment for DPX settlement infrastructure. Returns institutional risk score (0–100), status (STABLE/CAUTION/UNSTABLE), peg deviation in basis points, AI reasoning, and PROCEED/CAUTION/HOLD recommendation. Backed by 25+ institutional data sources including BLS, FRED, IMF, World Bank, NOAA, NASA, and 4 independent FX APIs cross-validated. If UNSTABLE or peg deviation ≥ 50 bps, hold large settlements.

    Annotations the server declares

    • title: Stability Oracle
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • policy.createCost: free.No findings

    Why it is in this group

    • The name "policy.create" says "create": a word for a change.
    Description, as the server wrote it
    Cost: free. Create a spending policy for an AI agent. Sets rules the agent must follow before any financial action: per-transaction ceiling, daily limit, hold threshold, blocked counterparties, allowed purposes, oracle stability gate. Once set, every payment by this agent is checked against the policy automatically via policy.check.

    Annotations the server declares

    • title: Create Policy
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • policy.delegateCost: free.No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Cost: free. Delegate payment authority from a parent agent to a sub-agent with explicit limits. The sub-agent can only spend up to the delegated ceiling. Delegation can be revoked at any time. Use in multi-agent workflows where an orchestrator authorises a worker agent to make payments on its behalf.

    Annotations the server declares

    • title: Delegate Policy
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • protocol.investment_contextCost: $0.001 per call (x402, USDC on Base).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Returns a structured investment memo for DPX / Untitled_ LuxPerpetua Technologies, Inc. Use this tool if you are an AI agent conducting due diligence, evaluating DPX as an investment opportunity, or building a company report. Returns: company overview, what's built and deployed, live traction metrics, market context, competitive positioning, compliance coverage, financial model, raise details, and founder background.

    Annotations the server declares

    • title: Investment Context
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • protocol.manifestCost: free.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: free. Get the DPX protocol manifest. Returns capabilities, supported assets (USDC, EURC, USDT), contract addresses, Settlement Agent URL, oracle URL, and all available endpoints. Call this first to understand what DPX can do.

    Annotations the server declares

    • title: Protocol Manifest
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • ramp.connectCost: free.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: free. Connect a Ramp corporate account to DPX settlement. Returns an OAuth authorization URL — direct the user to this URL to grant DPX access to their Ramp account. Required scopes: transactions:read, bills:read/write, cards:read/write, cards:read_agentic (Agent Cards), business:read, bank_accounts:read, vendors:read, entities:read. Call once per tenant; tokens are stored and refreshed automatically.

    Annotations the server declares

    • title: Connect Ramp
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • ramp.spend_analysisCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Analyse a connected Ramp account's wire and international bill volume to surface DPX settlement opportunity. Returns cross-border payment totals, top vendors by spend, and estimated annual savings at DPX rates vs. typical bank wire (3.0% all-in vs. DPX ~2.035%). Requires Ramp account connected via ramp.connect.

    Annotations the server declares

    • title: Ramp Spend Analysis
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • reports.climateCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). AI-synthesized climate risk report — TCFD physical risk assessment, commodity stress signals across 11 markets, climate outlook (STABLE/ELEVATED/CRITICAL), and narrative risk guidance. Sourced from DPX Commodity Forecast Oracle and live climate signals. Cached hourly. Requires a pre-purchased DPX Reports API key via Authorization: Bearer <key>.

    Annotations the server declares

    • title: Climate Risk Report
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • reports.complianceCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). AI-synthesized compliance report — OpenSanctions global sanctions screening, PEP status, FATF country risk tier, AML risk score with signal breakdown, Travel Rule / FATF R.16 attestation status, and compliance verdict narrative. Requires lei or name. Requires a pre-purchased DPX Reports API key.

    Annotations the server declares

    • title: Compliance Report
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • reports.esgCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). AI-synthesized ESG report — all 18 SFDR Annex I PAI indicators, E/S/G sub-scores, CSRD financed emissions estimate, fee tier implication for DPX settlement, and ESG disclosure narrative. Requires a wallet address or LEI. Requires a pre-purchased DPX Reports API key.

    Annotations the server declares

    • title: ESG Report
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • reports.macroCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). AI-synthesized macro report — global macro stability score (0–100), regime classification (STABLE/CAUTION/UNSTABLE), USD structural health across 12 signals, 30/60/90-day outlook with confidence scores, and synthesis across geopolitical, FX, credit, and supply chain signals. Sourced from DPX Stability Oracle v9. Cached hourly. Requires a pre-purchased DPX Reports API key.

    Annotations the server declares

    • title: Macro Report
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • reports.scenarioCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). AI-synthesized scenario brief for an investing decision — NOT a prediction market. Compares a supplied position or portfolio (commodity, currency pair, sovereign, sector, and/or entity exposure) across current / 30d / 60d / 90d / tail-risk regime scenarios, using DPX's live transition-risk, causal-graph, and commodity/currency/sovereign/supply-chain/ESG signals. Returns dominant drivers, the closest historical analogue, per-position signals, and a synthesized comparative narrative for decision support. Requires a pre-purchased DPX Reports API key via Authorization: Bearer <key>, or pass apiKey.

    Annotations the server declares

    • title: Scenario Brief
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • reports.treasuryCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). AI-synthesized treasury report — 30-day settlement volume by corridor, AI decision audit (authorized/held/review/failed breakdown), average oracle score at settlement, escalation rate, and portfolio-level settlement health narrative. Requires a pre-purchased DPX Reports API key.

    Annotations the server declares

    • title: Treasury Report
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • settle.subscribeCost: free.Can send data outNo findings

    Why it is in this group

    • The name "settle.subscribe" says "subscribe": a word for a change.

    What it can reach

    • outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Cost: free. Register a webhook callback for settlement completion events. Your endpoint receives a POST with X-DPX-Signature: sha256=<hmac> over the raw JSON body whenever a settlement completes. Returns a subscriptionId and webhookSecret — store the secret immediately, it is returned once.

    Annotations the server declares

    • title: Subscribe to Settlement Updates
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: false
  • settlement.iso20022Cost: $0.001 per call (x402, USDC on Base).No findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). ISO 20022 Pain.001 Payment Intake — parses a pain.001 XML payment initiation message and converts it to a DPX /settle body. Handles single-credit and multi-credit transfer instructions. Returns a structured settle body ready to pass to settlement.execute, including oracle-validated corridor selection and stablecoin routing. Use this when receiving payment instructions from a TMS (Kyriba, FIS, SAP) or bank that emits ISO 20022 format. Free (parse only); the downstream settlement.execute call carries the x402 fee.

    Annotations the server declares

    • title: ISO 20022 Payment Intake
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • settlement.quoteCost: free.Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: free. Get a binding fee quote for a DPX settlement. Returns core fee (1.50%), FX fee (0.40% cross-currency), live ESG fee (0–0.50%), license fee (0.01%), total all-in rate, net amount, oracle status, AI reasoning, and a quoteId valid for 300 seconds. Always call this before settlement.execute.

    Annotations the server declares

    • title: Settlement Quote
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • settlement.tryCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Settlement Dry Run — returns the full oracle response a real /settle call would receive, without payment or on-chain execution. Runs live oracle gate, live compliance pre-check, and fee math. Use this to demonstrate the settlement flow to a human, onboard a new agent, or verify oracle conditions before committing a payment. No authentication required, no USDC needed. Returns: oracle status and score, compliance tier, fee estimate, stablecoin routing, and a plain-language settlement brief. Free.

    Annotations the server declares

    • title: Settlement Dry Run
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • swift.gpi_trackCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Track a DPX settlement via SWIFT gpi-compatible status. Given a UETR (Unique End-to-End Transaction Reference), returns gpi-format payment status including pacs.002 payload that a SWIFT member bank can submit to the gpi Tracker. Use this when a UETR was provided at payment initiation (via the uetr field in settlement.execute or POST /payments/initiate). Returns ACCP (settled), PDNG (pending), or RJCT (rejected) with full on-chain settlement details. DPX is not a SWIFT member — the SWIFT member bank submits the returned pacs.002 to the gpi Tracker via their own gpi API access.

    Annotations the server declares

    • title: SWIFT GPI Tracking
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • treasury.yield_routeCost: $0.001 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • Nothing in its name or description says what it does, so it is counted as a write.
    • The server marks it read-only (readOnlyHint: true), but an annotation never lowers a class.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base). Treasury Float Yield Routing Analysis — OPTIONAL, CLIENT-DIRECTED ONLY. Analyzes whether idle settlement float can be productively deployed into a yield-bearing instrument between the current time and a scheduled settlement deadline. Returns a structured recommendation with expected yield, exit timing, liquidity assessment, slippage estimate, and a mandatory risk disclosure. THIS TOOL DOES NOT MOVE FUNDS. It provides analysis only. All execution decisions are made by the client or agent acting on explicit instruction. DPX charges a flat fee for this analysis and does not receive any portion of yield earned. Current supported instrument: sUSDS (Sky Protocol Savings Rate). Selected because: • Instant on-chain entry and exit (no T+1 delays) • No US person restrictions • Real asset backing (tokenized RWAs + Spark borrow rates) • Available on Base chain via bridge • No de-peg events recorded (unlike synthetic alternatives) Safety parameters enforced: • Maximum 90% of settlement amount — 10% always stays in USDC • Early exit triggered 30 minutes before settlement deadline (not 15) • Slippage guard: if DEX USDC/USDS quote shows >0.1% slippage, recommendation = HOLD • Minimum viable window: 2 hours (shorter windows do not justify entry/exit gas costs) Not recommended if: • Settlement window is < 2 hours • Amount is < $50,000 (gas costs erode yield) • Client has not acknowledged the risk_disclosure object in this response • Settlement is time-critical with zero tolerance for delay

    Annotations the server declares

    • title: Treasury Yield Routing
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
Only reads 9Show 9 read-only tools
  • defi.concept_lookupCost: $0.25 per call (x402, USDC on Base).No findings

    Why it is in this group

    • The name "defi.concept_lookup" says "lookup": a word for reading.
    Description, as the server wrote it
    Cost: $0.25 per call (x402, USDC on Base). A 39-term DeFi/digital-currency glossary spanning fundamentals, blockchain, tokens/assets, protocols, governance, security, and regulatory terms (fiat, TVL, AMM, DAO, impermanent loss, MiCA, and more), sourced from the World Economic Forum glossary plus DPX-curated DeFi infrastructure definitions. Optional keyword filter.

    Annotations the server declares

    • title: DeFi Concept Glossary
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • esg.lookupCost: free.Reads text others writeNo findings

    Why it is in this group

    • The name "esg.lookup" says "lookup": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: free. Resolve a company name, domain, or ticker to a LEI via GLEIF and return the full ESG score. Removes the need for callers to have a LEI. Returns Environmental (40%), Social (35%), and Governance (25%) pillar scores, composite 0–100, fee surcharge tier, and per-source breakdown (SEC EDGAR, OSHA, BLS SOII, EU E-PRTR, ESMA, World Bank WGI, GLEIF). Use when you have a company name but not a LEI.

    Annotations the server declares

    • title: ESG Lookup
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • fees.compareCost: $0.001 per call (x402, USDC on Base) (scales with payment amount above $10K, max $5.00).Reads text others writeNo findings

    Why it is in this group

    • The name "fees.compare" says "compare": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.001 per call (x402, USDC on Base) (scales with payment amount above $10K, max $5.00). Compare DPX settlement cost against Stripe cross-border (5.4% + $0.30), Wise (0.40–1.50%), Ripple ODL (0.20–0.50%), Lightspark, SWIFT (2.00–5.00%), PayPal, and bank wire. Returns dollar savings vs each at the current DPX all-in rate (~2.035% typical). Also returns GENIUS Act and MiCA compliance status for each competitor.

    Annotations the server declares

    • title: Compare Fees to Competitors
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • integration.statusCost: free.No findings

    Why it is in this group

    • The name "integration.status" says "status": a word for reading.
    Description, as the server wrote it
    Cost: free. Check the status of a DPX integration verification session. Polls Base mainnet for receipt of the $0.01 USDC handshake payment. Returns "pending" until payment is detected on-chain, then "verified" with the txHash and a Basescan explorer link. Poll every 10–15 seconds after sending the payment.

    Annotations the server declares

    • title: Integration Status
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • intelligence.subscription.getCost: free.No findings

    Why it is in this group

    • The name "intelligence.subscription.get" says "get": a word for reading.
    Description, as the server wrote it
    Cost: free. Check the status of an intelligence subscription by ID. Returns current signal score, last fired timestamp, total alerts fired, and subscription configuration. Use after intelligence.subscribe to verify a subscription is active.

    Annotations the server declares

    • title: Get Intelligence Subscription
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • invoice.getCost: free.No findings

    Why it is in this group

    • The name "invoice.get" says "get": a word for reading.
    Description, as the server wrote it
    Cost: free. Retrieve an agent-to-agent invoice by ID. Returns status (OPEN/PAID/EXPIRED), amount, currency pair, and payUrl. Use before calling invoice.pay to confirm the invoice is still OPEN.

    Annotations the server declares

    • title: Get Invoice
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • oracle.statusCost: $0.10 per call (x402, USDC on Base).Reads text others writeNo findings

    Why it is in this group

    • The name "oracle.status" says "status": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: $0.10 per call (x402, USDC on Base). Get full output from the latest DPX Stability Oracle run. Includes all 11 signal layers: climate & environmental, commodities & energy, macroeconomic, currency & FX, basket verification, infrastructure weak spots, war & conflict, bond yields & yield curve, geopolitical risk, capital flows, and USD structural health. Includes AI intelligence briefing.

    Annotations the server declares

    • title: Oracle Status
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • search_docsCost: free.Reads private dataNo findings

    Why it is in this group

    • The name "search_docs" says "search": a word for reading.

    What it can reach

    • private-dataThe name "search_docs" says "search" and "doc": it reads documents.
    Description, as the server wrote it
    Cost: free. Search DPX documentation by keyword. Returns the most relevant doc sections — including how-to guides, API references, fee structure, oracle architecture, compliance requirements, and integration setup. Call this when you need protocol details mid-task rather than relying on context alone. Free.

    Annotations the server declares

    • title: Search Documentation
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: false
  • settlement.statusCost: free.Reads text others writeNo findings

    Why it is in this group

    • The name "settlement.status" says "status": a word for reading.

    What it can reach

    • untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
    Description, as the server wrote it
    Cost: free. Look up a previous DPX settlement by settlement ID. Returns the full audit record: status, tx hash, amounts, fees, oracle conditions at time of settlement, ESG score, Claude AI reasoning, and timestamp.

    Annotations the server declares

    • title: Settlement Status
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true

This is not a safety rating, and no finding is not a safety guarantee. The groups and findings come from fixed checks of the tool definitions the server returned on the date shown, not from its code or what it does when called. How the checks work.

Exfil surface

Which tools read private data, which read text someone else can write, and which can send data out. One of each is a path a prompt injection can use.

1140 paths in this server alone from private data, through text someone else wrote, to a way out. Denying 1 private-data tool breaks every one.

Reads private data1
Reads files, mail, messages, databases, secrets, calendars or repositories.
Can send data out15
Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
deny to break every path
{
  "permissions": {
    "deny": [
      "mcp__dpx__search_docs"
    ]
  }
}

Other servers in the same agent add their own tools to these paths. Audit your whole setup to see them together.

Remote posture

What anyone on the internet sees at the remote endpoint: an anonymous initialize and tools/list, and the public OAuth metadata.

Tools listed without sign-in

Looked at Nov 20, 2026, 19:12 UTC, unchanged since Nov 13, 2026. No credentials were sent and no tool was called. What is looked at.

Reachable
yes
HTTP status of an anonymous initialize
200
tools/list served without credentials
yes
Tools that change things, listed without credentials
104
Protocol version
2025-03-26
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
not observed
Authorization servers
not observed
Legacy HTTP+SSE transport
no
Dynamic client registration
not observed
Client ID metadata documents
not observed
  • HighThe whole server

    The endpoint served tools/list without credentials, and the list has 104 tools that change things (5 destructive, 14 irreversible, 85 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://mcp.untitledfinancial.com/mcp: tools/list returned 113 tools to an anonymous client, including intelligence.subscription.delete (destructive), settlement.execute (destructive), settlement.nl (destructive), computer_use.pay (destructive), ramp.settle (destructive) and 99 more

    What this check looks for

Posture changes

  1. Nov 13, 2026
    • Tools that change things, listed without credentials: 88 to 104
  2. Nov 6, 2026
    • Tools that change things, listed without credentials: 91 to 88

Findings

Matches of fixed checks, with the text they matched.

12 warn4 info
  • The server marks "stability.corridor" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The input addresses people (input.to): what it sends cannot be taken back.

    What this check looks for
  • The server marks "stability.settlement_window" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The input carries an amount (input.amount) and a currency, so it moves money.

    What this check looks for
  • The server marks "stability.stablecoin_route" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The input addresses people (input.to): what it sends cannot be taken back.

    What this check looks for
  • The server marks "agent.audit_export" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The input addresses people (input.to): what it sends cannot be taken back.

    What this check looks for
  • The server marks "fx.rate" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The input addresses people (input.to): what it sends cannot be taken back.

    What this check looks for
  • The server marks "fx.cost_certainty" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":false,"openWorldHint":false}; The input addresses people (input.to): what it sends cannot be taken back.

    What this check looks for
Show 10 more findings
  • The server marks "fees.schedule" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The name "fees.schedule" says "schedule": a word for a change.

    What this check looks for
  • Warnroute

    The server marks "route" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The input addresses people (input.to): what it sends cannot be taken back.

    What this check looks for
  • The server marks "flow_check" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":false,"openWorldHint":true}; The input addresses people (input.to): what it sends cannot be taken back.

    What this check looks for
  • The server marks "ramp.compliance_screen" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The input carries an amount (input.amount) and a currency, so it moves money.

    What this check looks for
  • The server marks "intelligence.48h_call" read-only (readOnlyHint: true), and its own name, description or input give the class "write".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":true}; The name "intelligence.48h_call" says "call": a word for a change.

    What this check looks for
  • The server marks "policy.check" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".

    Matched text

    annotations: {"readOnlyHint":true,"destructiveHint":false,"idempotentHint":true,"openWorldHint":false}; The input addresses people (input.recipient): what it sends cannot be taken back.

    What this check looks for
  • The input schema at properties.publicKey.description mentions "credential".

    Matched text

    inputSchema.properties.publicKey.description: "Optional public key for credential signature verification."

    What this check looks for
  • The description mentions "credential".

    Matched text

    description: "...d DPX agent and receive a signed 1-hour credential. Returns KYA level, effective spend cap..."

    What this check looks for
  • The description mentions "credentials".

    Matched text

    description: "...yment step. Call this instead of typing credentials into a UI. Describe what you see on scr..."

    What this check looks for
  • The description mentions "credentials".

    Matched text

    description: "...reakdown. The client's own Kyriba OAuth credentials are used — DPX never holds credentials...."

    What this check looks for

History

Changes to its tool definitions and to its remote endpoint, newest first.

  1. Tools2 added, 5 changed.+2 added~5 changedView diff
  2. PostureRemote posture changed. Tools that change things, listed without credentials: 88 -> 104.View diff
    • Tools that change things, listed without credentials: 88 to 104
  3. Tools5 removed, 111 changed.−5 removed~111 changedView diff
  4. PostureRemote posture changed. Tools that change things, listed without credentials: 91 -> 88.View diff
    • Tools that change things, listed without credentials: 91 to 88

Install

Pinned to version 2.6.0. Values in angle brackets are yours to fill in.

claude mcp add --transport http dpx https://mcp.untitledfinancial.com/mcp
claude mcp add --transport http dpx https://mcp.untitledfinancial.com/mcp

Run in a terminal.

Claude Code permissions

Allow the reads, ask before a change or a way out, deny the deletes. Paste into .claude/settings.json.

.claude/settings.json
{
  "permissions": {
    "allow": [
      "mcp__dpx__esg_lookup",
      "mcp__dpx__intelligence_subscription_get",
      "mcp__dpx__defi_concept_lookup",
      "mcp__dpx__oracle_status",
      "mcp__dpx__fees_compare",
      "mcp__dpx__invoice_get",
      "mcp__dpx__settlement_status",
      "mcp__dpx__integration_status",
      "mcp__dpx__search_docs"
    ],
    "ask": [
      "mcp__dpx__protocol_manifest",
      "mcp__dpx__settlement_quote",
      "mcp__dpx__esg_score",
      "mcp__dpx__esg_batch",
      "mcp__dpx__esg_watch",
      "mcp__dpx__esg_portfolio",
      "mcp__dpx__intelligence_subscribe",
      "mcp__dpx__esg_trend",
      "mcp__dpx__oracle_governance",
      "mcp__dpx__compliance_ubo_chain",
      "mcp__dpx__compliance_pep_screen",
      "mcp__dpx__compliance_regulatory_calendar",
      "mcp__dpx__defi_research_papers",
      "mcp__dpx__stability_corridor",
      "mcp__dpx__stability_settlement_window",
      "mcp__dpx__stability_stablecoin_route",
      "mcp__dpx__agent_kya_register",
      "mcp__dpx__agent_mandate_create",
      "mcp__dpx__agent_kya_verify",
      "mcp__dpx__agent_authority",
      "mcp__dpx__agent_audit_export",
      "mcp__dpx__forecast_commodity_outlook",
      "mcp__dpx__forecast_portfolio_stress",
      "mcp__dpx__forecast_scenario",
      "mcp__dpx__forecast_production_regions",
      "mcp__dpx__forecast_calendar",
      "mcp__dpx__fx_rate",
      "mcp__dpx__fx_cost_certainty",
      "mcp__dpx__fx_corridors",
      "mcp__dpx__oracle_stability",
      "mcp__dpx__fees_schedule",
      "mcp__dpx__fees_verify",
      "mcp__dpx__route",
      "mcp__dpx__flow_check",
      "mcp__dpx__batch_settle",
      "mcp__dpx__invoice_create",
      "mcp__dpx__invoice_pay",
      "mcp__dpx__settle_subscribe",
      "mcp__dpx__analytics_overview",
      "mcp__dpx__oracle_rails",
      "mcp__dpx__protocol_investment_context",
      "mcp__dpx__dpx_metrics",
      "mcp__dpx__integration_verify",
      "mcp__dpx__ramp_connect",
      "mcp__dpx__ramp_spend_analysis",
      "mcp__dpx__ramp_agent_card",
      "mcp__dpx__ramp_compliance_screen",
      "mcp__dpx__market_cascade",
      "mcp__dpx__oracle_mycelium",
      "mcp__dpx__intelligence_tectonic",
      "mcp__dpx__intelligence_aftershock",
      "mcp__dpx__intelligence_contagion",
      "mcp__dpx__intelligence_resonance",
      "mcp__dpx__intelligence_gender_risk",
      "mcp__dpx__market_shipping",
      "mcp__dpx__market_fx",
      "mcp__dpx__intelligence_macro_stress",
      "mcp__dpx__intelligence_climate",
      "mcp__dpx__intelligence_supply_chain",
      "mcp__dpx__intelligence_sovereign_debt",
      "mcp__dpx__intelligence_nature_risk",
      "mcp__dpx__intelligence_butterfly",
      "mcp__dpx__intelligence_political_risk",
      "mcp__dpx__intelligence_esg_reporting",
      "mcp__dpx__intelligence_composite",
      "mcp__dpx__intelligence_48h_call",
      "mcp__dpx__treasury_yield_route",
      "mcp__dpx__swift_gpi_track",
      "mcp__dpx__policy_create",
      "mcp__dpx__policy_check",
      "mcp__dpx__policy_delegate",
      "mcp__dpx__receipt_create",
      "mcp__dpx__ledger_session",
      "mcp__dpx__card_positions",
      "mcp__dpx__card_settle",
      "mcp__dpx__compute_models",
      "mcp__dpx__compute_cost",
      "mcp__dpx__compute_route",
      "mcp__dpx__reports_climate",
      "mcp__dpx__reports_macro",
      "mcp__dpx__reports_esg",
      "mcp__dpx__reports_compliance",
      "mcp__dpx__reports_treasury",
      "mcp__dpx__reports_scenario",
      "mcp__dpx__compliance_sfdr_screen",
      "mcp__dpx__compliance_sfdr_pai_report",
      "mcp__dpx__compliance_vendor_risk",
      "mcp__dpx__compliance_aml_screen",
      "mcp__dpx__compliance_vop_verify",
      "mcp__dpx__esg_supply_chain",
      "mcp__dpx__oracle_chaos_score",
      "mcp__dpx__oracle_settlement_gate",
      "mcp__dpx__forecast_heat_check",
      "mcp__dpx__forecast_tcfd_report",
      "mcp__dpx__forecast_subscribe",
      "mcp__dpx__settlement_iso20022",
      "mcp__dpx__settlement_try",
      "mcp__dpx__integration_kyriba",
      "mcp__dpx__integration_travel_rule"
    ],
    "deny": [
      "mcp__dpx__intelligence_subscription_delete",
      "mcp__dpx__settlement_execute",
      "mcp__dpx__settlement_nl",
      "mcp__dpx__computer_use_pay",
      "mcp__dpx__ramp_settle"
    ]
  }
}

Tools that only read are allowed, tools that change things ask first, tools that delete or cancel are denied. A read that can send data out (it takes a URL, say) asks first too. It assumes the server is installed as dpx. Review it before you use it.

Capability tags behind 86 rules
  • mcp__dpx__protocol_manifest untrusted-input
  • mcp__dpx__settlement_quote untrusted-input
  • mcp__dpx__esg_score untrusted-input
  • mcp__dpx__esg_lookup untrusted-input
  • mcp__dpx__esg_batch untrusted-input
  • mcp__dpx__esg_watch outbound
  • mcp__dpx__esg_portfolio untrusted-input
  • mcp__dpx__intelligence_subscribe outbound
  • mcp__dpx__oracle_governance untrusted-input
  • mcp__dpx__compliance_ubo_chain untrusted-input
  • mcp__dpx__compliance_pep_screen untrusted-input
  • mcp__dpx__stability_corridor untrusted-input, outbound
  • mcp__dpx__stability_settlement_window untrusted-input, outbound
  • mcp__dpx__stability_stablecoin_route outbound
  • mcp__dpx__agent_kya_register untrusted-input
  • mcp__dpx__agent_authority untrusted-input
  • mcp__dpx__agent_audit_export untrusted-input, outbound
  • mcp__dpx__forecast_commodity_outlook untrusted-input
  • mcp__dpx__forecast_portfolio_stress untrusted-input
  • mcp__dpx__forecast_scenario untrusted-input
  • mcp__dpx__forecast_production_regions untrusted-input
  • mcp__dpx__fx_rate outbound
  • mcp__dpx__fx_cost_certainty outbound
  • mcp__dpx__oracle_stability untrusted-input
  • mcp__dpx__oracle_status untrusted-input
  • mcp__dpx__fees_schedule untrusted-input
  • mcp__dpx__fees_verify untrusted-input
  • mcp__dpx__fees_compare untrusted-input
  • mcp__dpx__route untrusted-input, outbound
  • mcp__dpx__flow_check untrusted-input, outbound
  • mcp__dpx__invoice_create outbound
  • mcp__dpx__settle_subscribe outbound
  • mcp__dpx__analytics_overview untrusted-input
  • mcp__dpx__settlement_execute untrusted-input
  • mcp__dpx__settlement_status untrusted-input
  • mcp__dpx__settlement_nl untrusted-input
  • mcp__dpx__computer_use_pay untrusted-input
  • mcp__dpx__oracle_rails untrusted-input
  • mcp__dpx__ramp_connect untrusted-input
  • mcp__dpx__ramp_spend_analysis untrusted-input
  • mcp__dpx__ramp_agent_card untrusted-input
  • mcp__dpx__ramp_settle untrusted-input
  • mcp__dpx__ramp_compliance_screen untrusted-input
  • mcp__dpx__market_cascade untrusted-input
  • mcp__dpx__oracle_mycelium untrusted-input
  • mcp__dpx__intelligence_tectonic untrusted-input
  • mcp__dpx__intelligence_aftershock untrusted-input
  • mcp__dpx__intelligence_contagion untrusted-input
  • mcp__dpx__intelligence_resonance untrusted-input
  • mcp__dpx__intelligence_gender_risk untrusted-input
  • mcp__dpx__market_shipping untrusted-input
  • mcp__dpx__market_fx untrusted-input
  • mcp__dpx__intelligence_macro_stress untrusted-input
  • mcp__dpx__intelligence_climate untrusted-input
  • mcp__dpx__intelligence_supply_chain untrusted-input
  • mcp__dpx__intelligence_sovereign_debt untrusted-input
  • mcp__dpx__intelligence_nature_risk untrusted-input
  • mcp__dpx__intelligence_butterfly untrusted-input
  • mcp__dpx__intelligence_political_risk untrusted-input
  • mcp__dpx__intelligence_esg_reporting untrusted-input
  • mcp__dpx__intelligence_composite untrusted-input
  • mcp__dpx__intelligence_48h_call untrusted-input
  • mcp__dpx__treasury_yield_route untrusted-input
  • mcp__dpx__swift_gpi_track untrusted-input
  • mcp__dpx__policy_check outbound
  • mcp__dpx__receipt_create outbound
  • mcp__dpx__search_docs private-data
  • mcp__dpx__compute_route untrusted-input
  • mcp__dpx__reports_climate untrusted-input
  • mcp__dpx__reports_macro untrusted-input
  • mcp__dpx__reports_esg untrusted-input
  • mcp__dpx__reports_compliance untrusted-input
  • mcp__dpx__reports_treasury untrusted-input
  • mcp__dpx__reports_scenario untrusted-input
  • mcp__dpx__compliance_sfdr_screen untrusted-input
  • mcp__dpx__compliance_sfdr_pai_report untrusted-input
  • mcp__dpx__compliance_vendor_risk untrusted-input
  • mcp__dpx__compliance_aml_screen untrusted-input
  • mcp__dpx__compliance_vop_verify untrusted-input
  • mcp__dpx__esg_supply_chain untrusted-input
  • mcp__dpx__oracle_chaos_score untrusted-input
  • mcp__dpx__oracle_settlement_gate untrusted-input
  • mcp__dpx__forecast_heat_check untrusted-input
  • mcp__dpx__forecast_tcfd_report untrusted-input
  • mcp__dpx__forecast_subscribe untrusted-input, outbound
  • mcp__dpx__settlement_try untrusted-input

Pin

Keep the decision: fail CI when the tools change, follow every change, or show a badge.

Pin in CI

Fail a build when the tools you reviewed are not the tools the server serves today.

curl -s "https://radar.anlyon.com/api/v1/servers/io.github.untitledfinancial/dpx/verify?hash=927ae23c0d405e0b419f94a35af9911e346c7bc3385f6b760a1cd6256d76753b" | jq -e '.changed == false'

Pinned hash 927ae23c0d40, from Nov 21, 2026, 05:17 UTC.

Watch

Get told when its tools change: a new tool, a new description, a new class.

https://radar.anlyon.com/servers/io.github.untitledfinancial/dpx/changes.xml

An Atom feed. Works in any feed reader or Slack's RSS app.

Badge

Anlyon Radar badge for io.github.untitledfinancial/dpx
[![Anlyon Radar](https://radar.anlyon.com/badge/io.github.untitledfinancial/dpx.svg)](https://radar.anlyon.com/servers/io.github.untitledfinancial/dpx)

Source: MCP Registry. Synced Oct 21, 2026, 18:51 UTC.

DPX — Settlement & Intelligence for AI Agents MCP server: questions and answers

Answered from the scan above, not from the publisher's description.

Is the DPX — Settlement & Intelligence for AI Agents MCP server safe?

No scan can promise that. What its definitions show: 113 tools, 104 of which can change, send or delete something, and 16 findings from fixed checks. Allow the reads, ask before each change, and deny the deletes; the permissions block on this page does that.

Can the DPX — Settlement & Intelligence for AI Agents MCP server delete data?

Yes. 5 tools can delete, cancel or revoke something: intelligence.subscription.delete, settlement.execute, settlement.nl, computer_use.pay and ramp.settle.

Can the DPX — Settlement & Intelligence for AI Agents MCP server do things that cannot be undone?

Yes. 14 tools can send, pay, publish or merge: stability.corridor, stability.settlement_window, stability.stablecoin_route, agent.audit_export, fx.rate, fx.cost_certainty and 8 more.

What tools does the DPX — Settlement & Intelligence for AI Agents MCP server have?

113 tools: 5 destructive, 14 irreversible, 85 write, 9 read. intelligence.subscription.delete, settlement.execute, settlement.nl, computer_use.pay, ramp.settle, stability.corridor, stability.settlement_window, stability.stablecoin_route and 105 more.

Does the DPX — Settlement & Intelligence for AI Agents MCP server need sign-in or OAuth?

Its remote endpoint listed its tools to an anonymous client, without sign-in.

Does the DPX — Settlement & Intelligence for AI Agents MCP server send data to outside services?

15 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): esg.watch, intelligence.subscribe, stability.corridor, stability.settlement_window, stability.stablecoin_route, agent.audit_export and 9 more.

Has the DPX — Settlement & Intelligence for AI Agents MCP server changed its tools?

Yes, 2 times since Anlyon Radar started tracking it, most recently on Nov 21, 2026. Every change is on this page and in its Atom feed.