Skip to content

NullClear Pre-Execution Firewall & Arbiter MCP server

io.github.Nullclear-Protocol/nullclearNullclear-Protocolstreamable-httpv1.6.0Scanned

7-chain EVM/Solana 1-call Pre-Sign Firewall, Guarded RPC, Proxy/ABI Inspector & x402 Escrow

14 tools: 2 do something that can't be undone, 7 change things and 5 only read.

Tools: 0 delete, 2 irreversible, 7 write, 5 read.
  • Reads private data · 1
  • Reads text others write · 14
  • Can send data out · 5
  • 7 findings

Tools listed without sign-in. Remote posture

Anlyon alternative

9 of these tools change something. Run those as governed actions.

Declare what an action changes, and Anlyon puts it under a shared limit, binds every approval to the exact request, and grades the outcome. It works on any HTTPS API you can describe as an action.

Anlyon Radar does not run, proxy or govern this MCP server. A tool your own code calls directly is not routed through Anlyon and is not governed by it.

Import an OpenAPI spec →

What the NullClear Pre-Execution Firewall & Arbiter MCP server can do

The NullClear Pre-Execution Firewall & Arbiter MCP server is published by the GitHub account Nullclear-Protocol. It has 14 tools. It can do things that cannot be undone, such as sending, paying or publishing (nullclear_simulate_transaction and nullclear_preflight_guard) and create or change data (7 tools: nullclear_scan_token, nullclear_decode_calldata, nullclear_audit_paid_token, nullclear_score_wallet and 3 more), but cannot delete data. 5 tools only read.

1 tool reads private data (nullclear_discover_apis); 14 tools bring in text other people can write (nullclear_scan_token, nullclear_inspect_contract, nullclear_decode_calldata and 11 more); 5 tools can send data out of your system (nullclear_simulate_transaction, nullclear_read_markdown, nullclear_preflight_guard and 2 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 7 findings (base64-blob and annotation-conflict). Each one comes with the text it matched.

Its remote endpoint lists its tools without sign-in, and its tools changed 3 times since Anlyon Radar first saw them.

Last scanned Nov 19, 2026, version 1.6.0.

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.

14 tools

Can't be undone 2Claude Code: ask
  • nullclear_preflight_guardAll-in-one 1-call Pre-Sign Firewall across Base, Ethereum, Arbitrum, Optimism, Polygon, BSC, and Solana (10 free calls/day/IP).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 "nullclear_preflight_guard" 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
    All-in-one 1-call Pre-Sign Firewall across Base, Ethereum, Arbitrum, Optimism, Polygon, BSC, and Solana (10 free calls/day/IP). In one pass, recursively unwraps Multicall3/Uniswap Universal Router/Safe calldata or EIP-712 Permits or Solana wire txs, runs stateful eth_simulateV1 / Solana simulateTransaction with asset balance diffs, and automatically scans all touched token contracts for honeypots, taxes, and Token-2022 permanent delegates. Returns one deterministic recommendedAgentAction (EXECUTE_SAFE / CAUTION_HIGH_RISK / ABORT_*).

    Annotations the server declares

    • title: 1-Call Multi-Chain Pre-Sign Firewall (Free)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • nullclear_simulate_transactionExecute stateful pre-flight simulation across EVM (eth_simulateV1 with stateOverrides, multi-tx bundles, and Transfer/Approval net asset diffs; fallback eth_call + eth_estimateGas)…Reads text others writeCan send data out1 finding

    Why it is in this group

    • The description's first verb is "execute": a word for a change.
    • 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 "nullclear_simulate_transaction" 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 description's first verb is "execute": a word for a change.

      What this check looks for
    Description, as the server wrote it
    Execute stateful pre-flight simulation across EVM (eth_simulateV1 with stateOverrides, multi-tx bundles, and Transfer/Approval net asset diffs; fallback eth_call + eth_estimateGas) and Solana (simulateTransaction + wire instruction audit) BEFORE signing (10 free calls/day/IP).

    Annotations the server declares

    • title: Stateful EVM & Solana Pre-Flight Simulator (Free)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
Changes things 7Claude Code: ask
  • nullclear_audit_paid_tokenRun NullClear's x402 Bazaar paid oracle for Base Mainnet (8453) ERC-20 tokens returning a cryptographically signed clearing receipt (10 free trial calls/day, then $0.005 USDC).Reads text others write2 findings

    Why it is in this group

    • The description's first verb is "run": 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 input schema at properties.tokenAddress.default contains a 42-character base64 string.

      Matched text

      inputSchema.properties.tokenAddress.default: "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913"

      What this check looks for
    • Warn

      The server marks "nullclear_audit_paid_token" 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 description's first verb is "run": a word for a change.

      What this check looks for
    Description, as the server wrote it
    Run NullClear's x402 Bazaar paid oracle for Base Mainnet (8453) ERC-20 tokens returning a cryptographically signed clearing receipt (10 free trial calls/day, then $0.005 USDC). Use this ONLY when you specifically require a Base Bazaar EIP-712/keccak256 settlement receipt alongside token LP and bytecode verification. For free (10/day) multi-chain Base/ETH/Solana token & proxy scanning, use nullclear_scan_token instead.

    Annotations the server declares

    • title: Base Token x402 Bazaar Attested Audit ($0.005)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • nullclear_decode_calldataDecode EVM calldata (recursively unwrapping Multicall3, Uniswap Universal Router, Safe execTransaction), EIP-712 typed-data signatures (Permit, Permit2, Seaport), Solana wire trans…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
    Decode EVM calldata (recursively unwrapping Multicall3, Uniswap Universal Router, Safe execTransaction), EIP-712 typed-data signatures (Permit, Permit2, Seaport), Solana wire transactions (Base64/Base58), custom Solidity revert bytes, or mined tx hashes with verified-router spender trust classification (10 free calls/day/IP).

    Annotations the server declares

    • title: Recursive Calldata, EIP-712, Solana Wire & Revert Decoder (Free)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • nullclear_execute_escrowExecute a paid x402 API request protected by NullClear deferred non-custodial EIP-3009 escrow, custom JSON Schema/value invariants, multi-seller failover routing, and outbound spen…Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "nullclear_execute_escrow" says "execute": 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.targetUrl): a request to a URL can carry data to whoever runs it.
    Description, as the server wrote it
    Execute a paid x402 API request protected by NullClear deferred non-custodial EIP-3009 escrow, custom JSON Schema/value invariants, multi-seller failover routing, and outbound spend caps (10 free trials/day, then requires configured wallet key). Settles upstream payment ONLY when all schema, numeric, staleness, and refusal checks pass; otherwise voids authorization ($0 lost). Use this instead of nullclear_fetch_guarded whenever calling third-party x402 paid APIs with strict output contracts or fallback sellers.

    Annotations the server declares

    • title: Deferred x402 Escrow Call with Schema & Failover
    • readOnlyHint: false
    • destructiveHint: false
    • idempotentHint: false
    • openWorldHint: true
  • nullclear_fetch_guardedExecute a zero-schema Auto-Guarded HTTP GET/POST request through NullClear's firewall against HTTP 5xx crashes, empty JSON bodies, Cloudflare/soft-paywall traps, and LLM refusal st…Reads text others writeCan send data out1 finding

    Why it is in this group

    • The name "nullclear_fetch_guarded" says "fetch": a word for reading.
    • The description's first verb is "execute": 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.
    • untrusted-inputThe description's first sentence says "execute" and "http": 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.

    Findings

    • Warn

      The server marks "nullclear_fetch_guarded" 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 "nullclear_fetch_guarded" says "fetch": a word for reading.

      What this check looks for
    Description, as the server wrote it
    Execute a zero-schema Auto-Guarded HTTP GET/POST request through NullClear's firewall against HTTP 5xx crashes, empty JSON bodies, Cloudflare/soft-paywall traps, and LLM refusal strings (10 free trials/day, then $0.005 USDC via Bazaar). Returns verified payload and executionReceipt, or voids settlement on failure. Do NOT use for plain HTML docs-to-Markdown conversion (use nullclear_read_markdown) or when custom JSON Schema/invariants/failover are needed (use nullclear_execute_escrow).

    Annotations the server declares

    • title: Auto-Guarded JSON/API Safe HTTP Fetch ($0.005)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • nullclear_scan_tokenPerform a free multi-chain (Base, Ethereum, Arbitrum, Optimism, Polygon, BSC, Solana) token security, proxy resolution, Token-2022 extension/mint/freeze authority audit, honeypot, …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.
    • 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 input schema at properties.address.default contains a 42-character base64 string.

      Matched text

      inputSchema.properties.address.default: "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913"

      What this check looks for
    Description, as the server wrote it
    Perform a free multi-chain (Base, Ethereum, Arbitrum, Optimism, Polygon, BSC, Solana) token security, proxy resolution, Token-2022 extension/mint/freeze authority audit, honeypot, and buy/sell tax audit (10 free calls/day/IP, zero API key; then $0.005 USDC). Returns score, verdict, recommendedAgentAction, proxy implementation, tax percentages, LP USD, and provenance block number (<400 tokens). Do NOT use for low-level 4-byte selector disassembly on non-token contracts (use nullclear_inspect_contract).

    Annotations the server declares

    • title: 7-Chain Token Security, Token-2022 & Proxy Scanner (Free)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • nullclear_score_walletProfile a raw 0x EVM wallet address on Base Mainnet (8453) for OFAC sanctions exposure, on-chain counterparty wash-trade concentration, ETH/USDC balances, transaction count, and ER…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
    Profile a raw 0x EVM wallet address on Base Mainnet (8453) for OFAC sanctions exposure, on-chain counterparty wash-trade concentration, ETH/USDC balances, transaction count, and ERC-8004 ownership (10 free trials/day, then $0.005 USDC). Returns riskScore (0-100), riskLevel, and ERC-8004 registration status. Do NOT pass a numeric ERC-8004 agent ID here — use nullclear_check_reputation for numeric agentId trust audits.

    Annotations the server declares

    • title: 0x Wallet Address Sanctions & Sybil Profiler ($0.005)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • nullclear_verify_mandateVerify an autonomous agent's Know Your Agent (KYA) EIP-712 corporate liability mandate, jurisdiction, expiration status, and whether a proposed USD transaction amount complies with…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
    Verify an autonomous agent's Know Your Agent (KYA) EIP-712 corporate liability mandate, jurisdiction, expiration status, and whether a proposed USD transaction amount complies with corporate per-transaction caps. Returns approved boolean, corporateEntity, liabilityCoverageUsd, maxTxLimitUsd, and expiration/limit flags. Use this specifically for corporate authorization checks — not for on-chain Sybil scoring (use nullclear_check_reputation).

    Annotations the server declares

    • title: Verify Corporate KYA Liability Mandate & Caps
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
Only reads 5Show 5 read-only tools
  • nullclear_check_reputationRetrieve an ERC-8004 registered agent's composite trust score (0-100), letter grade (A-F), historical clearing count, and Sybil wash-trading penalty flags by numeric agentId.Reads text others writeNo findings

    Why it is in this group

    • The description's first verb is "retrieve": 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
    Retrieve an ERC-8004 registered agent's composite trust score (0-100), letter grade (A-F), historical clearing count, and Sybil wash-trading penalty flags by numeric agentId. Returns arbiter-signed reputation metrics and sybilFlags. Use this ONLY for registered ERC-8004 agent IDs (e.g. '124595'); for raw 0x wallet address OFAC/balance profiling use nullclear_score_wallet, and for legal corporate liability limits use nullclear_verify_mandate.

    Annotations the server declares

    • title: Check ERC-8004 AgentId Trust & Sybil Score
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • nullclear_discover_apisSearch NullClear's audited x402 and ERC-8004 endpoint directory by keyword, category, minimum uptime, and maximum latency.Reads private dataReads text others writeNo findings

    Why it is in this group

    • The description's first verb is "search": a word for reading.

    What it can reach

    • private-dataThe description's first sentence says "search" and "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
    Search NullClear's audited x402 and ERC-8004 endpoint directory by keyword, category, minimum uptime, and maximum latency. Use this before paying external x402 sellers to find reliable APIs and retrieve their NullClear Zero-Loss Proxy URL (`https://api.nullclear.dev/https://...`). Returns matched endpoints with priceAtomic, uptimeScore, latencyMs, and proxy URL. Do NOT use to execute a call (pass the returned URL to nullclear_execute_escrow).

    Annotations the server declares

    • title: Search Audited x402 & ERC-8004 API Directory
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • nullclear_fetch_spot_priceQuery cross-exchange spot prices (Coinbase + Binance) for major crypto pairs with deterministic positive-price invariant verification and signed attestation (10 free trials/day, th…Reads text others writeNo findings

    Why it is in this group

    • The name "nullclear_fetch_spot_price" says "fetch": 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
    Query cross-exchange spot prices (Coinbase + Binance) for major crypto pairs with deterministic positive-price invariant verification and signed attestation (10 free trials/day, then $0.01 USDC). Returns priceUsd, source, invariantCheck ('PASSED'), and execution receipt. Use this whenever a verified non-hallucinated crypto spot price feed is needed.

    Annotations the server declares

    • title: Spread-Verified Crypto Spot Price Oracle ($0.01)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • nullclear_inspect_contractDisassemble deployed EVM bytecode across Base, Ethereum, Arbitrum, Optimism, Polygon, or BSC to resolve EIP-1967/1822/1167/Beacon/Diamond/Safe proxies, query Sourcify v2 verified A…Reads text others write1 finding

    Why it is in this group

    • The name "nullclear_inspect_contract" says "inspect": a word for reading.

    What it can reach

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

    Findings

    • Warn

      The input schema at properties.address.default contains a 42-character base64 string.

      Matched text

      inputSchema.properties.address.default: "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913"

      What this check looks for
    Description, as the server wrote it
    Disassemble deployed EVM bytecode across Base, Ethereum, Arbitrum, Optimism, Polygon, or BSC to resolve EIP-1967/1822/1167/Beacon/Diamond/Safe proxies, query Sourcify v2 verified ABIs, batch-resolve PUSH4 selectors via OpenChain, and analyze control-flow opcode risks (tx.origin, mutable DELEGATECALL, unguarded SELFDESTRUCT) without an Etherscan key (10 free calls/day/IP).

    Annotations the server declares

    • title: EVM Proxy, Sourcify ABI & Bytecode Inspector (Free)
    • readOnlyHint: true
    • destructiveHint: false
    • idempotentHint: true
    • openWorldHint: true
  • nullclear_read_markdownFetch and convert any public web documentation URL or HTML article into clean, stripped Markdown (<400 tokens) with a SHA-256 content hash.Reads text others writeCan send data outNo findings

    Why it is in this group

    • The name "nullclear_read_markdown" 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 description's first sentence says "fetch" and "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
    Fetch and convert any public web documentation URL or HTML article into clean, stripped Markdown (<400 tokens) with a SHA-256 content hash. Use this when reading documentation, blog posts, or web pages while minimizing context-window noise (10 free calls/day/IP). Returns markdown body and provenance hash. Do NOT use for JSON REST APIs or schema-enforced x402 data endpoints needing LLM-refusal/5xx protection (use nullclear_fetch_guarded or nullclear_execute_escrow).

    Annotations the server declares

    • title: HTML Documentation to Clean Markdown Reader (Free)
    • 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.

70 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 out5
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__nullclear__nullclear_discover_apis"
    ]
  }
}

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 17, 2026, 15:19 UTC, unchanged since Nov 17, 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
8
Protocol version
2024-11-05
WWW-Authenticate on 401
not observed
resource_metadata in WWW-Authenticate
not observed
RFC 9728 protected resource metadata
https://api.nullclear.dev/.well-known/oauth-protected-resource/mcp
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 8 tools that change things (1 irreversible, 7 write). Whether a call also works without credentials was not tested: tools are never called.

    Matched text

    https://api.nullclear.dev/mcp: tools/list returned 13 tools to an anonymous client, including nullclear_simulate_transaction (irreversible), nullclear_scan_token (write), nullclear_decode_calldata (write), nullclear_audit_paid_token (write), nullclear_score_wallet (write) and 3 more

    What this check looks for

Findings

Matches of fixed checks, with the text they matched.

7 warn
  • The input schema at properties.address.default contains a 42-character base64 string.

    Matched text

    inputSchema.properties.address.default: "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913"

    What this check looks for
  • The input schema at properties.address.default contains a 42-character base64 string.

    Matched text

    inputSchema.properties.address.default: "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913"

    What this check looks for
  • The server marks "nullclear_simulate_transaction" 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 description's first verb is "execute": a word for a change.

    What this check looks for
  • The server marks "nullclear_preflight_guard" 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 input schema at properties.tokenAddress.default contains a 42-character base64 string.

    Matched text

    inputSchema.properties.tokenAddress.default: "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913"

    What this check looks for
  • The server marks "nullclear_audit_paid_token" 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 description's first verb is "run": a word for a change.

    What this check looks for
Show 1 more finding
  • The server marks "nullclear_fetch_guarded" 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 "nullclear_fetch_guarded" says "fetch": a word for reading.

    What this check looks for

History

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

  1. Tools1 added, 6 changed.+1 added~6 changedView diff
  2. Tools12 added, 12 removed, 1 changed.+12 added−12 removed~1 changedView diff
  3. Tools3 changed.~3 changedView diff
    • Changed:nullclear_safe_fetchChanges thingsdescriptionannotations
    • Changed:nullclear_token_security_auditChanges thingsdescriptionannotations
    • Changed:nullclear_wallet_risk_scoreChanges thingsdescriptionannotations

Install

Pinned to version 1.6.0. Values in angle brackets are yours to fill in.

claude mcp add --transport http nullclear https://api.nullclear.dev/mcp
claude mcp add --transport http nullclear https://api.nullclear.dev/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__nullclear__nullclear_inspect_contract",
      "mcp__nullclear__nullclear_discover_apis",
      "mcp__nullclear__nullclear_fetch_spot_price",
      "mcp__nullclear__nullclear_check_reputation"
    ],
    "ask": [
      "mcp__nullclear__nullclear_scan_token",
      "mcp__nullclear__nullclear_decode_calldata",
      "mcp__nullclear__nullclear_simulate_transaction",
      "mcp__nullclear__nullclear_read_markdown",
      "mcp__nullclear__nullclear_preflight_guard",
      "mcp__nullclear__nullclear_audit_paid_token",
      "mcp__nullclear__nullclear_score_wallet",
      "mcp__nullclear__nullclear_fetch_guarded",
      "mcp__nullclear__nullclear_execute_escrow",
      "mcp__nullclear__nullclear_verify_mandate"
    ],
    "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 nullclear. Review it before you use it.

Capability tags behind 14 rules
  • mcp__nullclear__nullclear_scan_token untrusted-input
  • mcp__nullclear__nullclear_inspect_contract untrusted-input
  • mcp__nullclear__nullclear_decode_calldata untrusted-input
  • mcp__nullclear__nullclear_simulate_transaction untrusted-input, outbound
  • mcp__nullclear__nullclear_read_markdown untrusted-input, outbound
  • mcp__nullclear__nullclear_preflight_guard untrusted-input, outbound
  • mcp__nullclear__nullclear_discover_apis private-data, untrusted-input
  • mcp__nullclear__nullclear_audit_paid_token untrusted-input
  • mcp__nullclear__nullclear_score_wallet untrusted-input
  • mcp__nullclear__nullclear_fetch_spot_price untrusted-input
  • mcp__nullclear__nullclear_fetch_guarded untrusted-input, outbound
  • mcp__nullclear__nullclear_execute_escrow untrusted-input, outbound
  • mcp__nullclear__nullclear_check_reputation untrusted-input
  • mcp__nullclear__nullclear_verify_mandate 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.Nullclear-Protocol/nullclear/verify?hash=88f3e4c09d63f5ab4ca73f75ba6cf80433abf83d31ab793d5ae8509192fe3439" | jq -e '.changed == false'

Pinned hash 88f3e4c09d63, from Nov 19, 2026, 13:18 UTC.

Watch

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

https://radar.anlyon.com/servers/io.github.Nullclear-Protocol/nullclear/changes.xml

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

Badge

Anlyon Radar badge for io.github.Nullclear-Protocol/nullclear
[![Anlyon Radar](https://radar.anlyon.com/badge/io.github.Nullclear-Protocol/nullclear.svg)](https://radar.anlyon.com/servers/io.github.Nullclear-Protocol/nullclear)

Source: MCP Registry · repository. Synced Nov 19, 2026, 13:17 UTC.

NullClear Pre-Execution Firewall & Arbiter MCP server: questions and answers

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

Is the NullClear Pre-Execution Firewall & Arbiter MCP server safe?

No scan can promise that. What its definitions show: 14 tools, 9 of which can change, send or delete something, and 7 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 NullClear Pre-Execution Firewall & Arbiter MCP server delete data?

No tool is classed as destructive: going by their definitions, none of them deletes, cancels or revokes anything.

Can the NullClear Pre-Execution Firewall & Arbiter MCP server do things that cannot be undone?

Yes. 2 tools can send, pay, publish or merge: nullclear_simulate_transaction and nullclear_preflight_guard.

What tools does the NullClear Pre-Execution Firewall & Arbiter MCP server have?

14 tools: 2 irreversible, 7 write, 5 read. nullclear_simulate_transaction, nullclear_preflight_guard, nullclear_scan_token, nullclear_decode_calldata, nullclear_audit_paid_token, nullclear_score_wallet, nullclear_fetch_guarded, nullclear_execute_escrow and 6 more.

Does the NullClear Pre-Execution Firewall & Arbiter MCP server need sign-in or OAuth?

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

Does the NullClear Pre-Execution Firewall & Arbiter MCP server send data to outside services?

5 tools can carry data out (it sends, posts or publishes, or takes a recipient, channel or URL): nullclear_simulate_transaction, nullclear_read_markdown, nullclear_preflight_guard, nullclear_fetch_guarded and nullclear_execute_escrow.

Has the NullClear Pre-Execution Firewall & Arbiter MCP server changed its tools?

Yes, 3 times since Anlyon Radar started tracking it, most recently on Nov 19, 2026. Every change is on this page and in its Atom feed.