TrustGraph releases AI Agent output policy control tool
Understand this story
TrustGraph has released a policy-as-code toolset that uses a single policy file to specify who can read content derived from each data source, which agents can use it, and where action outputs are sent, along with allow, needs approval, or deny decisions and the reasons for them. The toolset includes a CLI, gateway, serve, and a Python SDK, and requires Rust 1.90 or later. TrustGraph says the policy CLI and trustgraph serve have passed security review; the gateway and Python integration are new components, and two rounds of review found and fixed real vulnerabilities, with a third round still in progress. The company recommends trying it out but not yet using it with real data.
AI overview based on source reports · Updated 51 min ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AITrustGraph – ACLs that follow the data through your AI agents
TrustGraph is a policy-as-code tool that records what each AI agent output was built from and decides whether that output may be sent to a given audience, returning allow, needs approval or deny with a reason. It ships a Rust CLI, an HTTP/MCP service, a gateway that sits between agents and MCP servers, and a Python SDK with LangChain, OpenAI Agents SDK and Claude Agent SDK integrations. The author states the policy CLI and trustgraph serve have been through security review, while the gateway and Python integrations are new and should not yet be put in front of real data.
Activity for this story
Not enough continuous observations to show a trend yet.