Audifact
Signed, hash-chained receipts of an AI agent's decisions with public proof links
- Pricing
- Freemium
- Platforms
- API
- Runs
- Remotely
- Version
- 0.1.1
- Listing updated
- 2 days ago
About Audifact
Audifact is a freemium MCP server for security. It runs as a hosted remote server, so there is nothing to install locally.
Audifact is built around one idea: log what an agent knew at the moment it decided, then attach the outcome later, in a ledger that cannot be edited or deleted. Every event is signed and hash-chained, and the agent gets back a proof link in the form audifact.io/e/{event_hash}. The plugin's README is clear that this is receipts, not full tracing, and suggests roughly 3 to 10 events per active day.
The hosted server at api.audifact.io/mcp exposes log_decision, log_outcome, get_proof, register_agent, list_agent_decisions and list_agent_outcomes. A bundled skill tells the agent to record consequential decisions, external actions and subagent delegations while redacting secrets, tokens, passwords and full message bodies. Public labels on proof links are off by default, and each agent logs under its own agent_name.
The key goes in a masked field (X-API-Key), and every event uses credits on the account that owns it. The wider product also offers tracers for LangChain, LlamaIndex, LangGraph, CrewAI and AutoGen, with a free tier and a demo mode that needs no key.
Install
Run in a terminal
claude mcp add --transport http audifact https://api.audifact.io/mcpAdd to Settings → Connectors → Add custom connector
https://api.audifact.io/mcpAdd to ~/.cursor/mcp.json
{
"mcpServers": {
"audifact": {
"url": "https://api.audifact.io/mcp"
}
}
}Add to .vscode/mcp.json
{
"servers": {
"audifact": {
"type": "http",
"url": "https://api.audifact.io/mcp"
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"audifact": {
"serverUrl": "https://api.audifact.io/mcp"
}
}
}Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"audifact": {
"type": "http",
"url": "https://api.audifact.io/mcp"
}
}
}Tools this server provides (6)
Taken from the project documentation; the running server may differ.
- log_decision
- Records a decision in the signed ledger
- log_outcome
- Attaches an outcome to a decision
- get_proof
- Returns proof for a logged event
- register_agent
- Registers an agent
- list_agent_decisions
- Lists an agent's decisions
- list_agent_outcomes
- Lists an agent's outcomes
What it does
- Signed decision log
- Each decision is written to an append-only, hash-chained ledger.
- Outcome linking
- Outcomes can be attached to earlier decisions once results exist.
- Public proof links
- Share a link per event, with optional public labels.
- Agent registration
- register_agent and per-agent names separate logs by agent.
- Redaction guidance
- The skill tells agents to avoid secrets and full message bodies.
- Plugin packaging
- Claude Code marketplace install plus an Agent Skills option.
Good for
- Prove what an agent decided and why during an audit
- Record subagent delegations for later review
- Share a proof link for a consequential automated action
- Review an agent's recent decisions and outcomes
Pricing
Freemium
Free tier; events consume account credits
Prices change. Check audifact.io before you buy.
Audifact FAQ
Can logged entries be removed?
No. The README says everything logged is signed, hash-chained and cannot be edited or deleted.
What do I need to connect?
An Audifact account and an API key (af_...) supplied in a secret field.
Does it cost anything?
Each event uses credits on the account that owns the key; the site mentions a free tier to start.
How often should an agent log?
The README targets about 3 to 10 events per active day.
Is Audifact free?
Audifact has a free plan, with paid plans for more usage or features.
How do I add Audifact to Claude Code?
Run this command in your terminal: claude mcp add --transport http audifact https://api.audifact.io/mcp
What are the best alternatives to Audifact?
The closest alternatives to Audifact are REA, Claude Code Ultimate Guide, Agent Roast, Midplane and Leash.
Alternatives to Audifact

REA
MCP serverCLI and MCP server that lets coding agents reverse engineer apps and binaries
Runs locally- Open source
- 464K GitHub stars
- Developer tools

Claude Code Ultimate Guide
MCP serverMCP server for searching the Claude Code Ultimate Guide from any MCP client
Runs locally- Open source
- 6.5K GitHub stars
- AI models and LLM tooling
Agent Roast
MCP serverOffline review that flags agent changes which weaken tests, assertions or CI
Runs locally- Open source
- 4 Hacker News points
- Developer tools

Midplane
MCP serverPostgres gateway for AI agents with table grants, column masking and approval for writes
Runs locally- Open source
- 5 GitHub stars
- Databases

Leash
MCP serverReview what your AI touched, approve risky actions and undo supported changes
- 1 tool
- Remote
- Free
- 1 GitHub star
- Security
Zero-log MCP server with 49 tools for AI guardrails, secret scanning and EU compliance
- 49 tools
- Remote
- Freemium
- Security