Tool Call Audit Trail

protect-mcp — Policy Enforcement + Signed Receipts

Cryptographic governance for every Claude Code tool call. Each invocation is evaluated against a Cedar policy and produces an Ed25519-signed receipt that anyone can verify offline.

Overview

Claude Code runs powerful tools: Bash, Edit, Write, WebFetch. By default there is no audit trail, no policy enforcement, and no way to prove what was decided after the fact. protect-mcp closes all three gaps:

• Cedar policies (AWS's open authorization engine) evaluate every tool call before execution. Cedar deny is authoritative.
• Ed25519 receipts record each decision with its inputs, the policy that governed it, and the outcome. Receipts are hash-chained.
• Offline verification via npx @veritasacta/verify. No server, no account, no trust in the operator.

Problem

AI agents make decisions that affect money, safety, and rights. The Claude Code session log records what happened, but the log is:

• Mutable — anyone with access can edit it
• Unsigned — there is no way to prove integrity
• Operator-bound — verification requires trusting whoever holds the log

For compliance contexts (finance, healthcare, regulated research), this is not sufficient. You need tamper-evident evidence that can be verified by third parties without trusting you.

Solution

Add protect-mcp to your Claude Code project:

# 1. Install the plugin (adds hooks + skill to your project)
claude plugin install wshobson/agents/protect-mcp

# 2. Create ./protect.cedar (see below). The plugin installs the hooks.

# 3. Start the receipt-signing server (runs locally, no external calls)
npx protect-mcp@latest serve --enforce

# 4. Use Claude Code normally. Every tool call is now policy-evaluated
#    and produces a signed receipt in ./receipts/

Hook Configuration

Installing the plugin adds both hooks from hooks/hooks.json. Each hook runs a script bundled with the plugin:

{
  "hooks": {
    "PreToolUse": [
      {
        "matcher": ".*",
        "hooks": [
          { "type": "command", "command": "\"${CLAUDE_PLUGIN_ROOT}\"/hooks/evaluate.sh" }
        ]
      }
    ],
    "PostToolUse": [
      {
        "matcher": ".*",
        "hooks": [
          { "type": "command", "command": "\"${CLAUDE_PLUGIN_ROOT}\"/hooks/sign.sh" }
        ]
      }
    ]
  }
}

Claude Code passes the hook event to the command as JSON on stdin and does not set TOOL_NAME or TOOL_INPUT variables. evaluate.sh reads tool_name and tool_input from that payload and passes them to protect-mcp as flags; sign.sh reads tool_name only, because the 0.7.4 signer records nothing else. Set PROTECT_MCP_POLICY, PROTECT_MCP_RECEIPTS, and PROTECT_MCP_KEY to change the default paths. When the policy file is missing, the PreToolUse hook prints a warning to stderr and allows the call.

What each hook does

PreToolUse — Runs BEFORE the tool executes. Evaluates the tool call against your Cedar policy file. If Cedar returns deny, the hook exits with code 2 and Claude Code blocks the tool call entirely.

PostToolUse — Runs AFTER the tool completes. Signs a receipt containing the tool name, input hash, output hash, decision, policy digest, and timestamp. Writes the receipt to ./receipts/<timestamp>.json.

Cedar Policy File

Create ./protect.cedar at the project root:

// Allow read-only tools by default
permit (
    principal,
    action in [Action::"Read", Action::"Glob", Action::"Grep", Action::"WebFetch"],
    resource
);

// Require explicit allow for destructive tools
permit (
    principal,
    action == Action::"Bash",
    resource
) when {
    // Allow safe commands only
    context.command_pattern in ["git", "npm", "ls", "cat", "echo", "pwd", "test"]
};

// Never allow recursive deletion
forbid (
    principal,
    action == Action::"Bash",
    resource
) when {
    context.command_pattern == "rm -rf"
};

// Require confirmation for writes outside the project
forbid (
    principal,
    action in [Action::"Edit", Action::"Write"],
    resource
) when {
    context.path_starts_with != "."
};

Verification

Verify a single receipt:

npx @veritasacta/verify receipts/2026-04-15T10-30-00Z.json
# Exit 0 = valid
# Exit 1 = tampered
# Exit 2 = malformed

Verify the entire chain:

npx @veritasacta/verify receipts/*.json

Use the plugin's slash commands from within Claude Code:

/verify-receipt receipts/latest.json
/audit-chain ./receipts/ --last 20

Receipt Format

Each receipt is a JSON file with this structure:

{
  "receipt_id": "rec_8f92a3b1",
  "receipt_version": "1.0",
  "issuer_id": "claude-code-protect-mcp",
  "event_time": "2026-04-15T10:30:00.000Z",
  "tool_name": "Bash",
  "input_hash": "sha256:a3f8...",
  "decision": "allow",
  "policy_id": "autoresearch-safe",
  "policy_digest": "sha256:b7e2...",
  "parent_receipt_id": "rec_3d1ab7c2",
  "public_key": "4437ca56815c0516...",
  "signature": "4cde814b7889e987..."
}

• Ed25519 signatures (RFC 8032)
• JCS canonicalization (RFC 8785) before signing
• Hash-chained to the previous receipt via parent_receipt_id
• Offline verifiable — no network call, no vendor lookup

Why This Matters

Before	After
"Trust me, the agent only read files"	Cryptographically provable: every Read logged and signed
"The log shows it happened"	The receipt proves it happened, and no one can edit it
"You'd have to audit our system"	Anyone can verify every receipt offline
"Logs might be different by now"	Ed25519 signatures lock the record at signing time

Standards

• Ed25519 — RFC 8032 (digital signatures)
• JCS — RFC 8785 (deterministic JSON canonicalization)
• Cedar — AWS's open authorization policy language
• IETF draft — draft-farley-acta-signed-receipts

Related

• npm: protect-mcp
• Verify CLI: @veritasacta/verify
• Source: github.com/ScopeBlind/scopeblind-gateway
• Protocol: veritasacta.com
• Integrations: Microsoft Agent Governance Toolkit (PR #667), AWS cedar-policy/cedar-for-agents (PR #64)