Agent Utils (x402) MCP server
dev.workers.revenue-agent-jw.agent-utils-x402-production/agent-utilsagent-utils-x402-production.revenue-agent-jw.workers.devstreamable-httpv1.14.0Scanned
Public records, OSV, PII redact, web reader, stablecoin freeze check for agents. x402 USDC
27 tools: 1 do something that can't be undone, 22 change things and 4 only read.
- Reads private data · 2
- Reads text others write · 26
- Can send data out · 3
- 3 findings
Tools listed without sign-in. Remote posture
With Anlyon
1 tool here does something that can't be undone. Govern it.
- Anlyon lists the tools and adopts none.
- Each tool you adopt gets approvals and receipts.
- Your agent never holds the key.
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 Agent Utils (x402) MCP server can do
The Agent Utils (x402) MCP server is published by agent-utils-x402-production.revenue-agent-jw.workers.dev, a domain the MCP Registry verified. It has 27 tools. It can do things that cannot be undone, such as sending, paying or publishing (us_business_days) and create or change data (22 tools: recall_check, sanctions_address_check, sanctions_name_screen, climate_day and 18 more), but cannot delete data. 4 tools only read.
2 tools read private data (federal_register_search and fda_ndc_lookup); 26 tools bring in text other people can write (recall_check, sanctions_address_check, sanctions_name_screen and 23 more); 3 tools can send data out of your system (us_business_days, x402_check and web_read). 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 3 findings (annotation-conflict). Each one comes with the text it matched.
Its remote endpoint lists its tools without sign-in, and its tools changed 10 times since Anlyon Radar first saw them.
Last scanned Nov 16, 2026, version 1.14.0.
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.
27 tools
Can't be undone 1Claude Code: ask
us_business_daysUS business days calculator: counts working days between two dates (inclusive), excluding weekends and observed US federal holidays per 5 U.S.C.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 "us_business_days" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
What this check looks forMatched text
annotations: {"readOnlyHint":true,"idempotentHint":true,"openWorldHint":true}; The input addresses people (input.to): what it sends cannot be taken back.
Description, as the server wrote it US business days calculator: counts working days between two dates (inclusive), excluding weekends and observed US federal holidays per 5 U.S.C. 6103 and OPM rules. Useful for deadlines, SLAs, shipping and payment due dates. Returns the count and lists the holidays skipped. Errors: 400 JSON {error} for invalid input. PAID: $0.002 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/calendar/us-business-days.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
Changes things 22Claude Code: ask
caller_avoidance_scoreCaller Avoidance Score (callers-avoid-0.1, outcome-only; honesty deferred): toxicity / avoid grade for a Crypto Twitter handle from public-call follower sim (post+5m, fees).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 Caller Avoidance Score (callers-avoid-0.1, outcome-only; honesty deferred): toxicity / avoid grade for a Crypto Twitter handle from public-call follower sim (post+5m, fees). Returns avoid_score 0-100, band, components with receipts (sim_mean, sim_win_rate, rug_rate, hit_2x_24h + Wilson CI), verdict (copy_buy_loses|insufficient_sample|mixed), and public post id links — not full post text. Free when n_scored<4 via GET /v0/callers/score. Not financial advice. PAID: $0.002 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/callers/score.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
climate_dayDaily high and low temperature for a US weather station from NWS: running provisional high/low over the station's climate day (midnight-midnight local standard time) and the offici…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 Daily high and low temperature for a US weather station from NWS: running provisional high/low over the station's climate day (midnight-midnight local standard time) and the official NWS Daily Climate Report (CLI) high/low once issued, flagged preliminary or final. Useful for temperature settlement, weather prediction markets and climate records. Optional temperature buckets (width/offset or explicit edges). Errors: 400 JSON {error} for invalid input. PAID: $0.002 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/weather/climate-day.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
clinical_trial_studyClinicalTrials.gov study lookup by NCT ID (official v2 API).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 ClinicalTrials.gov study lookup by NCT ID (official v2 API). Useful when an agent needs the facts of one trial. Returns titles, sponsor and collaborators, status and why stopped, key dates, phases, design, enrollment, conditions, interventions, primary outcomes, eligibility (ages, sex, criteria text), summary, site count, countries and up to 25 sites, plus results link, source URL and retrievedAt. Errors: 400 bad NCT ID, 404 not found. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/clinical-trials/study.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
email_auth_auditEmail authentication DNS audit for a domain or email address: SPF with recursive 10-lookup count and all qualifier, DMARC policy, pct and report addresses, DKIM selector probe, MTA…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 Email authentication DNS audit for a domain or email address: SPF with recursive 10-lookup count and all qualifier, DMARC policy, pct and report addresses, DKIM selector probe, MTA-STS with policy file, TLS-RPT, BIMI and MX. Useful for deliverability checks, Google/Yahoo bulk-sender readiness and spoofing risk. Returns grade A-F, score, spoofable flag, parsed records and plain-language issues. Errors: 400 invalid domain, 404 NXDOMAIN. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/dns/email-auth.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
fda_adverse_eventsFDA adverse event report counts (FAERS via openFDA) for a drug by generic or brand name, optionally filtered by reaction, date range or serious reports only.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 FDA adverse event report counts (FAERS via openFDA) for a drug by generic or brand name, optionally filtered by reaction, date range or serious reports only. Useful for drug safety signal checks and pharmacovigilance research. Returns total, serious, non-serious and death report counts and the top reported reactions, with source URL, FDA data date and retrievedAt. Reports do not prove causation. Errors: 400 invalid input. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/fda/adverse-events.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
fda_drug_labelFDA drug label (package insert, SPL) lookup via openFDA by brand or generic name, NDC or set ID; prefers the original packager's latest label.Reads text others write1 finding
Why it is in this group
- The name "fda_drug_label" says "label": 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 "fda_drug_label" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
What this check looks forMatched text
annotations: {"readOnlyHint":true,"idempotentHint":true,"openWorldHint":true}; The name "fda_drug_label" says "label": a word for a change.
Description, as the server wrote it FDA drug label (package insert, SPL) lookup via openFDA by brand or generic name, NDC or set ID; prefers the original packager's latest label. Useful for indications, dosing, boxed warnings, contraindications, interactions and adverse reactions. Returns label metadata (brand, generic, manufacturer, application number, NDCs, RxCUI) and selected sections trimmed to max_chars, plus DailyMed link and retrievedAt. Errors: 400 invalid input, 404 no label. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/fda/drug-label.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
fixture_settleKnown-answer x402 payment fixture for auditor and client self-tests.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 Known-answer x402 payment fixture for auditor and client self-tests. A paid call settles exactly at $0.001 USDC and returns the published known answer plus a signed receipt (payer, amount, tx, server time). Useful to prove a wallet and x402 client work end to end. Failed payments are not charged and return a machine-readable error. Spec: GET /v0/fixture/spec. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/fixture/settle.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
gas_estimateLive EVM gas fees on Ethereum, Base, Arbitrum, Optimism, Polygon, Avalanche: next base fee, slow/standard/fast tips (median of p10/p50/p90 over 20 blocks), EIP-1559 maxFee, and cos…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 Live EVM gas fees on Ethereum, Base, Arbitrum, Optimism, Polygon, Avalanche: next base fee, slow/standard/fast tips (median of p10/p50/p90 over 20 blocks), EIP-1559 maxFee, and cost of a native or ERC-20 transfer in native and USD (Chainlink), incl. OP-stack L1 data fee. Optional gas=units. Picks cheapest chain. 502 if a chain is unreadable (not charged). PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/gas/estimate.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
injection_scanPrompt-injection detection for untrusted text (web pages, emails, documents, tool output) before it reaches an LLM; a guardrail for agent pipelines.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 Prompt-injection detection for untrusted text (web pages, emails, documents, tool output) before it reaches an LLM; a guardrail for agent pipelines. Flags instruction overrides, role spoofing, secret/system-prompt exfiltration, image beacons, tool-call markup, hidden Unicode, hidden HTML and base64 payloads. Returns a 0-100 risk score, risk level and matched rules. Deterministic rules, no LLM. JSON body {text}, max 32 KB. Errors: 400 missing text, 413 too large. PAID: $0.002 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: POST /v1/text/injection-scan.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
osv_package_vulnsOpen-source package vulnerability check via OSV.dev for npm, PyPI, Go, crates.io, Maven, RubyGems, NuGet, Packagist and more, by ecosystem+name+version, purl, or up to 10 purls.Reads text others write1 finding
Why it is in this group
- The description's first verb is "open": 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 "osv_package_vulns" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
What this check looks forMatched text
annotations: {"readOnlyHint":true,"idempotentHint":true,"openWorldHint":true}; The description's first verb is "open": a word for a change.
Description, as the server wrote it Open-source package vulnerability check via OSV.dev for npm, PyPI, Go, crates.io, Maven, RubyGems, NuGet, Packagist and more, by ecosystem+name+version, purl, or up to 10 purls. Useful before installing or upgrading a dependency. Returns vulnerable flag, count, CVE/GHSA aliases, CVSS score and severity, fixed versions and a verified upgradeTo version. Errors: 400 invalid input, 502 OSV unavailable. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/osv/package.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
recall_checkProduct and vehicle recall lookup across US agencies in one call: NHTSA (cars, by VIN or make/model/year), CPSC (consumer products, by UPC or product name) and FDA enforcement repo…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 Product and vehicle recall lookup across US agencies in one call: NHTSA (cars, by VIN or make/model/year), CPSC (consumer products, by UPC or product name) and FDA enforcement reports (food, drugs, medical devices). Useful for safety checks before buying, selling or using a product or vehicle. Returns normalized recalls (number, date, hazard/reason, remedy, link), newest first, plus anyRecall. Errors: 400 JSON {error} for invalid input. PAID: $0.002 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/recalls/check.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
sanctions_address_checkOFAC sanctions screening for crypto wallet addresses: exact match against every digital currency address on the U.S.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 OFAC sanctions screening for crypto wallet addresses: exact match against every digital currency address on the U.S. Treasury OFAC SDN list (BTC, ETH, USDT, USDC, TRX, SOL, LTC, XMR and more). Useful for compliance and AML checks before paying a counterparty or an x402 payTo. Up to 20 comma-separated addresses; returns sanctioned true/false, SDN name, programs, official link and list freshness. Errors: 400 JSON {error} for invalid input. PAID: $0.002 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/sanctions/address.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
sanctions_name_screenOFAC sanctions name screening for KYC/AML/KYB: fuzzy match a person, company, vessel or aircraft name against every primary name and alias on the U.S.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 OFAC sanctions name screening for KYC/AML/KYB: fuzzy match a person, company, vessel or aircraft name against every primary name and alias on the U.S. Treasury OFAC SDN and Consolidated (non-SDN) lists. Accent, word-order and suffix tolerant. Up to 10 names (names=A|B). Returns verdict, score, list, programs, DOB, countries, official link. 400 invalid input. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/sanctions/name.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
sec_13f_holdingsSEC 13F-HR institutional holdings for a filer by CIK or ticker (latest quarter or a given accession), parsed from the EDGAR information table.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 SEC 13F-HR institutional holdings for a filer by CIK or ticker (latest quarter or a given accession), parsed from the EDGAR information table. Useful for tracking what funds like Berkshire Hathaway hold. Returns period, holdings ranked by value with issuer, CUSIP, class, shares, put/call and portfolio percent, totals checked against the cover page, recent 13F filings and source URLs. Tables over 1.5 MB return 422. Errors: 400, 404 no 13F. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/sec/13f.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
sec_8k_filingsSEC 8-K current reports from EDGAR: one company's 8-Ks by ticker or CIK, or the latest 8-Ks market-wide, optionally by item code (2.02 earnings, 5.02 officer changes, 1.01 agreemen…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 SEC 8-K current reports from EDGAR: one company's 8-Ks by ticker or CIK, or the latest 8-Ks market-wide, optionally by item code (2.02 earnings, 5.02 officer changes, 1.01 agreements) or full-text phrase. Useful for event monitoring and research. Returns company, CIK, tickers, filing date, accession number, decoded items and document and index URLs, with retrievedAt. Errors: 400 invalid input, 503 SEC rate limit busy. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/sec/8k.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
sec_filingsSEC EDGAR company filings and financials for a stock ticker or CIK in one call: latest 10-K, 10-Q, 8-K and other forms with dates, accession numbers, primary document and index URL…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 SEC EDGAR company filings and financials for a stock ticker or CIK in one call: latest 10-K, 10-Q, 8-K and other forms with dates, accession numbers, primary document and index URLs, plus optional XBRL facts (revenue, net income, shares outstanding). Useful for company research, due diligence and earnings monitoring. Errors: 400 JSON {error} for an unknown ticker; 503 while the shared SEC rate limit is busy. PAID: $0.002 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/sec/filings.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
solana_pre_promoPre-promo tape lookup for a Solana token when we hold a locked [-60m,0] window around its DexScreener paid-promo timestamp (specialty covered shelf).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 Pre-promo tape lookup for a Solana token when we hold a locked [-60m,0] window around its DexScreener paid-promo timestamp (specialty covered shelf). Returns raw aggregates: net buy USD, unique buyers, largest buyer share, and the same vs a matched-null mean from other covered tokens. Uncovered token: 404 with coverage:false, never fabricated. Free probe: GET /v0/solana/pre-promo-coverage. Not financial advice. PAID: $0.01 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/solana/pre-promo.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
stablecoin_blacklist_checkCan this wallet move USDC or USDT?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 Can this wallet move USDC or USDT? Live check of issuers' own on-chain blacklists: Circle USDC (Base, Ethereum, Arbitrum, Optimism, Polygon, Avalanche), EURC (Base, Ethereum), Tether USDT (Ethereum, Tron), plus OFAC SDN exact match. Up to 20 EVM 0x or Tron T addresses. Returns PASS/BLOCK, which token+chain is frozen, block numbers, contracts read. Fails closed: chain unreadable = 502, not charged. 400 invalid address. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/wallet/stablecoin-blacklist.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
text_pii_redactDeterministic PII detect-and-redact for untrusted text (tickets, chat logs, LLM prompts) before storage or model input.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 Deterministic PII detect-and-redact for untrusted text (tickets, chat logs, LLM prompts) before storage or model input. Finds email, phone, US SSN, credit card (Luhn), IP, IBAN (mod-97), US ABA routing, and secret prefixes (sk-, ghp_, AKIA…). Checksum-validated; no LLM/NER. Returns findings with offsets, placeholders, redacted copy and countsByType. Optional types= filter. POST JSON {text} max 32 KB (GET ?text= max 2,000). Errors: 400 bad input. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: POST /v1/text/pii-redact.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
us_federal_holidaysFREE.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 FREE. US federal holidays (observed dates) for a year, per 5 U.S.C. 6103. Useful to test the connection before paying.
Annotations the server declares
- readOnlyHint: true
- idempotentHint: true
us_weatherUS weather forecast, current conditions and severe-weather alerts for a latitude/longitude from the National Weather Service (NWS) in one call.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 US weather forecast, current conditions and severe-weather alerts for a latitude/longitude from the National Weather Service (NWS) in one call. Useful when an agent needs local temperature, a forecast or active alerts for a US location. Returns the latest station observation, next 12 hourly periods, 6-period forecast and active alerts as normalized JSON. Errors: 400 JSON {error} for invalid input. PAID: $0.002 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/weather/us.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
x402_checkx402 endpoint validator and payment-readiness checker, useful for sellers debugging a paywall or buyers vetting an endpoint before paying.Reads text others writeCan 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.
- 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 URL (input.url): a request to a URL can carry data to whoever runs it.
Description, as the server wrote it x402 endpoint validator and payment-readiness checker, useful for sellers debugging a paywall or buyers vetting an endpoint before paying. Fetches the URL unpaid, decodes the 402 challenge and grades network, USDC asset, payTo, price vs facilitator fees, EIP-712 domain, CORS and Bazaar metadata. Returns a 0-100 score and a fix list. Never pays the target. Errors: 400 JSON {error} for invalid input. PAID: $0.002 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/x402/check.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
Only reads 4Show 4 read-only tools
clinical_trials_searchClinicalTrials.gov study search (official v2 API) by condition, intervention/drug, sponsor, location, keyword, recruitment status and phase.Reads text others writeNo findings
Why it is in this group
- The name "clinical_trials_search" says "search": 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 ClinicalTrials.gov study search (official v2 API) by condition, intervention/drug, sponsor, location, keyword, recruitment status and phase. Useful for trial finding, pipeline and competitive research. Returns total count, NCT IDs, titles, status, phases, conditions, interventions, sponsor, enrollment, key dates, has-results flag and study links, with paging token, source URL and retrievedAt. Errors: 400 invalid input. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/clinical-trials/search.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
fda_ndc_lookupFDA National Drug Code (NDC) directory lookup via openFDA by product or package NDC (dashed, or 10/11-digit billing format) or by drug name.Reads private dataReads text others writeNo findings
Why it is in this group
- The name "fda_ndc_lookup" says "lookup": a word for reading.
What it can reach
- private-dataThe description's first sentence says "directory": it reads files.
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Description, as the server wrote it FDA National Drug Code (NDC) directory lookup via openFDA by product or package NDC (dashed, or 10/11-digit billing format) or by drug name. Useful for pharmacy, claims and inventory data. Returns product NDC, brand and generic names, labeler, dosage form, route, active ingredients and strengths, marketing category and dates, application number, DEA schedule and packages, with DailyMed link and retrievedAt. Errors: 400 invalid input, 404 no match. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/fda/ndc.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
federal_register_searchFederal Register search for US rules, proposed rules, notices and presidential documents by keyword, agency, type and publication date, or one document by number.Reads private dataReads text others writeNo findings
Why it is in this group
- The name "federal_register_search" says "search": a word for reading.
What it can reach
- private-dataThe description's first sentence says "document": it reads documents.
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
Description, as the server wrote it Federal Register search for US rules, proposed rules, notices and presidential documents by keyword, agency, type and publication date, or one document by number. Useful for regulatory monitoring and compliance research. Returns title, type, action, abstract, agencies, effective and comment-close dates, docket IDs, RINs, CFR parts, citation and official HTML/PDF links, with source URL and retrievedAt. Errors: 400 invalid input, 404 unknown document. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/federal-register/search.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: true
web_readWeb page reader for agents: fetch one public URL and get clean Markdown or text of the main content plus title, description, canonical, language, dates, headings and absolute links…Reads text others writeCan send data outNo findings
Why it is in this group
- The name "web_read" says "read": a word for reading.
What it can reach
- untrusted-inputThe server marks it openWorldHint: true: it reaches things outside the server's own data.
- untrusted-inputThe name "web_read" says "read" and "web": it reads web pages, which anyone can write.
- untrusted-inputThe description's first sentence says "web": it reads web pages, which anyone can write.
- untrusted-inputThe input takes a URL (input.url), and the tool reads it: it reads web pages, which anyone can write.
- outboundThe input takes a URL (input.url): a request to a URL can carry data to whoever runs it.
Claude Code: ask, because it is tagged outbound.
Description, as the server wrote it Web page reader for agents: fetch one public URL and get clean Markdown or text of the main content plus title, description, canonical, language, dates, headings and absolute links. Strips nav, scripts and cookie banners; follows redirects; handles JSON/text. Honors robots.txt and never bypasses logins or bot blocks. Errors are not charged: 400 bad URL, 422 robots/404/blocked, 502 unreachable. PAID: $0.001 USDC per call on Base via x402 (MCP transport: retry with params._meta["x402/payment"]). Same tool over plain HTTP: GET /v1/web/read.
Annotations the server declares
- readOnlyHint: true
- openWorldHint: true
- idempotentHint: 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.
156 paths in this server alone from private data, through text someone else wrote, to a way out. Denying 2 private-data tools breaks every one.
- Reads private data2
- Reads files, mail, messages, databases, secrets, calendars or repositories.
- Reads text others write26
- Brings in text someone else can write: web pages, inbound email, messages, issues and comments.
- Can send data out3
- Can carry data to people or systems: it sends, posts or publishes, or takes a recipient, channel or URL.
{
"permissions": {
"deny": [
"mcp__agent-utils__fda_ndc_lookup",
"mcp__agent-utils__federal_register_search"
]
}
}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.
Looked at Nov 21, 2026, 19:24 UTC, unchanged since Nov 21, 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
- 23
- Protocol version
- 2025-06-18
- 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 23 tools that change things (1 irreversible, 22 write). Whether a call also works without credentials was not tested: tools are never called.
What this check looks forMatched text
https://agent-utils-x402-production.revenue-agent-jw.workers.dev/mcp: tools/list returned 27 tools to an anonymous client, including us_business_days (irreversible), recall_check (write), sanctions_address_check (write), sanctions_name_screen (write), climate_day (write) and 18 more
Posture changes
- Nov 21, 2026
- Tools that change things, listed without credentials: 22 to 23
- Nov 14, 2026
- Tools that change things, listed without credentials: 19 to 22
- Nov 7, 2026
- Tools that change things, listed without credentials: 10 to 19
- Oct 30, 2026
- Tools that change things, listed without credentials: 8 to 10
Findings
Matches of fixed checks, with the text they matched.
- Warnus_business_days
The server marks "us_business_days" read-only (readOnlyHint: true), and its own name, description or input give the class "irreversible".
What this check looks forMatched text
annotations: {"readOnlyHint":true,"idempotentHint":true,"openWorldHint":true}; The input addresses people (input.to): what it sends cannot be taken back.
- Warnfda_drug_label
The server marks "fda_drug_label" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
What this check looks forMatched text
annotations: {"readOnlyHint":true,"idempotentHint":true,"openWorldHint":true}; The name "fda_drug_label" says "label": a word for a change.
The server marks "osv_package_vulns" read-only (readOnlyHint: true), and its own name, description or input give the class "write".
What this check looks forMatched text
annotations: {"readOnlyHint":true,"idempotentHint":true,"openWorldHint":true}; The description's first verb is "open": a word for a change.
History
Changes to its tool definitions and to its remote endpoint, newest first.
PostureRemote posture changed. Tools that change things, listed without credentials: 22 -> 23.View diff
- Tools that change things, listed without credentials: 22 to 23
Tools1 added.+1 addedView diff
- Added:sanctions_name_screenChanges things
PostureRemote posture changed. Tools that change things, listed without credentials: 19 -> 22.View diff
- Tools that change things, listed without credentials: 19 to 22
Tools1 added, 1 changed.+1 added~1 changedView diff
- Added:gas_estimateChanges things
- Changed:text_pii_redactChanges thingsdescription
Tools3 changed.~3 changedView diff
- Changed:email_auth_auditChanges thingsdescription
- Changed:osv_package_vulnsChanges thingsdescription
- Changed:stablecoin_blacklist_checkChanges thingstitledescriptioninputSchema
Tools1 added, 8 changed.+1 added~8 changedView diff
- Added:stablecoin_blacklist_checkChanges things
- Changed:clinical_trial_studyChanges thingsdescription
- Changed:clinical_trials_searchOnly readsdescription
- Changed:fda_adverse_eventsChanges thingsdescription
- Changed:fda_drug_labelChanges thingsdescription
- Changed:fda_ndc_lookupOnly readsdescription
- Changed:federal_register_searchOnly readsdescription
- Changed:sec_13f_holdingsChanges thingsdescription
- Changed:sec_8k_filingsChanges thingsdescription
Tools1 added.+1 addedView diff
- Added:fixture_settleChanges things
Tools1 added, 1 changed.+1 added~1 changedView diff
- Added:web_readOnly reads
- Changed:solana_pre_promoChanges thingsdescription
PostureRemote posture changed. Tools that change things, listed without credentials: 10 -> 19.View diff
- Tools that change things, listed without credentials: 10 to 19
Tools1 added.+1 addedView diff
- Added:text_pii_redactChanges things
Install
Pinned to version 1.14.0. Values in angle brackets are yours to fill in.
claude mcp add --transport http agent-utils https://agent-utils-x402-production.revenue-agent-jw.workers.dev/mcpclaude mcp add --transport http agent-utils https://agent-utils-x402-production.revenue-agent-jw.workers.dev/mcpRun 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.
{
"permissions": {
"allow": [
"mcp__agent-utils__federal_register_search",
"mcp__agent-utils__clinical_trials_search",
"mcp__agent-utils__fda_ndc_lookup"
],
"ask": [
"mcp__agent-utils__recall_check",
"mcp__agent-utils__sanctions_address_check",
"mcp__agent-utils__sanctions_name_screen",
"mcp__agent-utils__climate_day",
"mcp__agent-utils__us_weather",
"mcp__agent-utils__sec_filings",
"mcp__agent-utils__injection_scan",
"mcp__agent-utils__us_business_days",
"mcp__agent-utils__x402_check",
"mcp__agent-utils__caller_avoidance_score",
"mcp__agent-utils__clinical_trial_study",
"mcp__agent-utils__fda_drug_label",
"mcp__agent-utils__fda_adverse_events",
"mcp__agent-utils__sec_8k_filings",
"mcp__agent-utils__sec_13f_holdings",
"mcp__agent-utils__email_auth_audit",
"mcp__agent-utils__osv_package_vulns",
"mcp__agent-utils__text_pii_redact",
"mcp__agent-utils__solana_pre_promo",
"mcp__agent-utils__web_read",
"mcp__agent-utils__fixture_settle",
"mcp__agent-utils__stablecoin_blacklist_check",
"mcp__agent-utils__gas_estimate",
"mcp__agent-utils__us_federal_holidays"
],
"deny": []
}
}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 agent-utils. Review it before you use it.
Capability tags behind 26 rules
- mcp__agent-utils__recall_check untrusted-input
- mcp__agent-utils__sanctions_address_check untrusted-input
- mcp__agent-utils__sanctions_name_screen untrusted-input
- mcp__agent-utils__climate_day untrusted-input
- mcp__agent-utils__us_weather untrusted-input
- mcp__agent-utils__sec_filings untrusted-input
- mcp__agent-utils__injection_scan untrusted-input
- mcp__agent-utils__us_business_days untrusted-input, outbound
- mcp__agent-utils__x402_check untrusted-input, outbound
- mcp__agent-utils__caller_avoidance_score untrusted-input
- mcp__agent-utils__federal_register_search private-data, untrusted-input
- mcp__agent-utils__clinical_trials_search untrusted-input
- mcp__agent-utils__clinical_trial_study untrusted-input
- mcp__agent-utils__fda_drug_label untrusted-input
- mcp__agent-utils__fda_adverse_events untrusted-input
- mcp__agent-utils__fda_ndc_lookup private-data, untrusted-input
- mcp__agent-utils__sec_8k_filings untrusted-input
- mcp__agent-utils__sec_13f_holdings untrusted-input
- mcp__agent-utils__email_auth_audit untrusted-input
- mcp__agent-utils__osv_package_vulns untrusted-input
- mcp__agent-utils__text_pii_redact untrusted-input
- mcp__agent-utils__solana_pre_promo untrusted-input
- mcp__agent-utils__web_read untrusted-input, outbound
- mcp__agent-utils__fixture_settle untrusted-input
- mcp__agent-utils__stablecoin_blacklist_check untrusted-input
- mcp__agent-utils__gas_estimate 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/dev.workers.revenue-agent-jw.agent-utils-x402-production/agent-utils/verify?hash=dd7de70901b2f35c2d4257c09a940283fd7cdf3df604a1bedc0a06682e214c94" | jq -e '.changed == false'Pinned hash dd7de70901b2, from Nov 16, 2026, 17:18 UTC.
Watch
Get told when its tools change: a new tool, a new description, a new class.
https://radar.anlyon.com/servers/dev.workers.revenue-agent-jw.agent-utils-x402-production/agent-utils/changes.xmlAn Atom feed. Works in any feed reader or Slack's RSS app.
Badge
[](https://radar.anlyon.com/servers/dev.workers.revenue-agent-jw.agent-utils-x402-production/agent-utils)Source: MCP Registry. Synced Nov 16, 2026, 17:17 UTC.
Agent Utils (x402) MCP server: questions and answers
Answered from the scan above, not from the publisher's description.
Is the Agent Utils (x402) MCP server safe?
- No scan can promise that. What its definitions show: 27 tools, 23 of which can change, send or delete something, and 3 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 Agent Utils (x402) MCP server delete data?
- No tool is classed as destructive: going by their definitions, none of them deletes, cancels or revokes anything.
Can the Agent Utils (x402) MCP server do things that cannot be undone?
- Yes. 1 tool can send, pay, publish or merge: us_business_days.
What tools does the Agent Utils (x402) MCP server have?
- 27 tools: 1 irreversible, 22 write, 4 read. us_business_days, recall_check, sanctions_address_check, sanctions_name_screen, climate_day, us_weather, sec_filings, injection_scan and 19 more.
Does the Agent Utils (x402) MCP server need sign-in or OAuth?
- Its remote endpoint listed its tools to an anonymous client, without sign-in.
Does the Agent Utils (x402) MCP server send data to outside services?
- 3 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): us_business_days, x402_check and web_read.
Has the Agent Utils (x402) MCP server changed its tools?
- Yes, 10 times since Anlyon Radar started tracking it, most recently on Nov 16, 2026. Every change is on this page and in its Atom feed.