Guardian: open-source authority layer pins human approval to each exact agent tool call
manofstyle04 · reddit · 2026-11-12
The author releases Guardian (Apache-2.0), an authority layer for AI agents: every tool call gets a deterministic ALLOW/ASK/DENY decision from your rules — no model involved.
Key design: approval covers exactly one call, not the session:
- On ASK, a human sees the actual tool, target, arguments and amount before deciding
- Approval is pinned to a canonical SHA-256 hash of the action; the agent claims a one-time signed token, expiring and scoped to that agent and adapter
- At execution, Guardian verifies the hash: any changed argument is refused, as are reused or expired tokens; SDK wrappers run the approved arguments, not whatever the caller passes
Guards against cases like approving "email the Q3 report to finance" and having the agent send something else, or twice. Ships as TypeScript/Python SDKs with LangChain, Vercel AI and OpenAI Agents wrappers, plus an MCP server.
More from coding & agent
- Hackathon project lets AI agents spend company money, but only within CFO-approved limits — kleffew94 · 2026-11-12
- In real orgs the model is rarely the problem — the context it gets is, says Salesforce engineer — msrivastav13 · 2026-11-12
- Salesforce Winter '27 adds AiAgentDefinition metadata for versioned agent deployments — msrivastav13 · 2026-11-12
- Solo dev builds zombie horde level with Tripo P2.0, Fable and new Kaigen engine — gdechichi · 2026-11-12
- United Airlines to Ban Personal AI Agents From Booking Flights for Customers — Polymarket · 2026-11-12
- Agentic Search explained: a 4-search loop costs about 5x one search — and queries leak to more companies — adnan_hashmi · 2026-11-12