Audit Ledger Mcp

shahidh68/audit-ledger-mcp

If you're shipping AI in healthcare, finance, or anywhere the EU AI Act or FCA model risk rules apply, this server gives you a one-line path from Claude Desktop or LangGraph to a tamper-evident audit trail. It exposes three tools: record_decision logs AI outputs with locally hashed PII, verify_decision checks integrity against S3 Object Lock storage, and list_decisions queries your tenant's history. The sandbox mode lets you test with zero config using npx, writing to a shared public ledger. For production, point it at your own deployed audit-ledger backend and get 7-year COMPLIANCE-mode retention in DynamoDB and S3. Personal data never leaves your machine unhashed, so you can log GDPR-sensitive decisions without transmitting raw PII to storage.

GitHub stars: 1
tools: 4
auth: Secret needed
transport: STDIO
env vars: 6
last updated: Jun 5, 2026

View on GitHub

Tools

4 tools

Public tool metadata for what this MCP can expose to an agent.

record_decisionRecord an AI decision to the ledger.

Record an AI decision to the ledger.

No parameter schema in public metadata yet.

verify_decisionTamper-check a stored record.

Tamper-check a stored record.

No parameter schema in public metadata yet.

verify_completenessDetect missing records.

Detect missing records.

No parameter schema in public metadata yet.

list_decisionsList recent decisions for the calling tenant.

List recent decisions for the calling tenant.

No parameter schema in public metadata yet.

Configuration

AUDIT_API_URL

Your deployed audit-ledger API endpoint (e.g. https://<api-id>.execute-api.<region>.amazonaws.com/prod). Leave unset to run in sandbox mode against the public shared tenant.

AUDIT_WRITE_KEYsecret

Tenant write key for your deployed ledger. Required if AUDIT_API_URL is set and you want to call record_decision.

AUDIT_READ_KEYsecret

Tenant read key for your deployed ledger. Required if AUDIT_API_URL is set and you want to call verify_decision or list_decisions.

AUDIT_TIMEOUT_MS

HTTP timeout in milliseconds. Defaults to 5000.

AUDIT_RETRY_ATTEMPTS

Number of retry attempts on transient failures. Defaults to 3.

AUDIT_HMAC_KEYsecret

Tenant secret used to HMAC-SHA256 PII and prompts locally before sending. Generate once per tenant (e.g. node -e "console.log(require('crypto').randomBytes(32).toString('hex'))") and keep it in your environment. The key never leaves your process. If unset, the MCP falls back to p

Readme

audit-ledger-mcp

Connect Claude, Cursor, LangGraph, or your own agent to the AI Audit Ledger. This MCP server gives an agent the tools to record, check, and list decisions in a tamper-evident log with one line of config.

It is built for teams that need a clear record of AI decisions: EU AI Act Article 12 logging, FCA SS1/23 model risk evidence, and GDPR data minimisation. Raw personal data is hashed locally before anything is sent, so the ledger only sees fingerprints.

View the full README on GitHub

Details

Repository: github.com/shahidh68/audit-ledger-mcp
Category: AI & LLM Tools
Package: audit-ledger-mcp
Transport: STDIO
Auth: Required
Registry: MCP registry, active

Related AI & LLM Tools MCP servers

View all

Windy Wordsneakyfree/windy-word-mcpControl Windy Word voice-to-text from any MCP agent — 60 tools across 12 categories.1stars

Synapbussynapbus/synapbusMCP-native agent-to-agent messaging hub with channels, DMs, and semantic search1stars

X402search Mcpx402-index/x402search-mcpSearch 13,000+ x402-enabled APIs from your AI agent. Powered by x402search.xyz1stars

Xenarch Mcpxenarch-ai/xenarch-mcpXenarch — x402 MCP server for AI agent payments. Non-custodial, USDC on Base L2.1stars

Gmemyksanjo/gmemPersistent, Solana-aware project memory for AI coding agents. Typed entities + git-aware diffs.1stars

Deeprepoabhishek2432001/deeprepoProductivity-boosting RAG engine for codebases with multi-provider AI support and semantic search.0stars