
Microsoft toolkit for policy enforcement, identity and audit of autonomous agents
- Open source
- Open-source agent
- 6.6K GitHub stars
- Security

Microsoft toolkit for policy enforcement, identity and audit of autonomous agents

Open-source gateway adding identity, policy and audit between AI agents and your systems
Checks RFC 9421 HTTP message signatures and 402 payment credentials, per call
Experimental agent wallet that seals agreements in a public register anchored in Bitcoin
Runs locallyPassive audit that checks whether an MCP endpoint exposes its tools without authentication
Runs locallySign agent actions into a tamper-evident log and verify any agent's log from an MCP client
Screens crypto wallets and names against US Treasury OFAC lists, pay per call
Pay-per-call security screening for AI agents: OFAC, SSRF, secrets and diffs
Remote MCP server for Curve Compliance's healthcare marketing and consent reports
Not checked yet
Read-only checks of an AI agent's Sigvara standing, score, stake and slashes
Runs locallyLet an AI agent register and pay for its own verifiable DID-LLM identity
Pay-per-call audit of a public URL: redirects, TLS, security headers and DNS records
Pay-per-call crypto token risk, web and developer data tools for agents, paid with x402
A safer MCP for Dokploy: agents deploy and fix apps without seeing keys or secrets
Not checked yetAsk Claude about on-site incidents, monthly counts and trends in SISA
Bulk domain lookup: tech stack, email provider, DNS, WHOIS age, SSL expiry, AI bot rules
Pay-per-call Bitcoin fees, transaction proofs, address risk checks and whale feeds
MCP wrapper that holds database writes until a second person approves them on their phone
Runs locallyCheck a store or product page for scam and dark-pattern signals before buying
Runs locallyPay-per-check on-chain risk verdicts for EVM tokens and wallets, built for agents
Real-time alerts for deepfake voices, impersonation and social engineering on live calls
Not checked yet
Open-source control plane that firewalls MCP tool calls and caps agent model spend

Agent Data Tools compliance: OFAC sanctions, counterparty, company, domain, email checks

Read-only payment preflight for AI agents: pay, review or do-not-pay before signing
Pay-per-call USDT payment checks and address risk scores for agents
Run an 88-test MCP spec compliance suite and get an A-F grade from CLI or agent
Runs locallyLocal input firewall that scans text, files and MCP metadata for prompt injection
Runs locallynpm registry intelligence, security audits and publish-side writes without the 2FA prompt
Runs locallyPay-per-proof service showing a hidden value met a threshold, without revealing it
Not checked yetScreen an x402 payment counterparty with allow, review or block before an agent pays
Runs locally