action-receipts

Agent Receipts / Obsigna plugin for OpenClaw — cryptographically signed audit trail for agent actions

Governed MCP server for IP cameras (VAPIX + ONVIF SOAP): fail-closed per-agent policy, hash-chained signed receipts on every call — including denials. First reference producer for the AAR spec.

Cryptographic verification for AI agent actions — ECDSA-secp256k1 + RFC 6979 signed Action Receipts (v0.1), multi-dimensional Trust Vector, capability tokens (JWT-shaped + Biscuit-style attenuation), UETA §10(b) undo. 29 MCP tools. A2A v1.0 + ERC-8004 format compatible. Receipt-batch Merkle roots anchored on Base mainnet. Starting with code.

Local-first Solana transaction policy firewall and tamper-evident receipt ledger for AI agents.

Zero-trust API firewall and security integrity layer for autonomous AI agents & Model Context Protocol (MCP) tool execution. Secure, TOCTOU-proof, fail-closed.

A standard for mapping AI agent action paths, browser receipts, execution logs, and monetization events into audit-ready royalty candidates.

TypeScript SDK for OnChainDiligence — agent payment lifecycle (preflight, execution, settlement observation, receipts) plus a typed client for paid Evidence Provider checks (sanctions, company) via x402.

An interactive audit, policy, and receipt flow for trustworthy AI agent actions.

Verifiable-AI MCP server: mint and verify offline action receipts, run reproducible GDPR checks. Local-first, no telemetry.

Agent Action Receipts — vendor-neutral, evidence-grade conformance profiles for AI agents acting on physical-security systems. Spec (CC BY 4.0) + free offline verifier (Apache-2.0).

OS Tag: open control plane for human-agent collaboration — governed, auditable, multi-agent work loops from Feishu/Lark, DingTalk, Slack threads or the CLI