Skip to content
Hot eventDeveloping

TrustGraph releases policy-as-code tool for AI agent data flows

1 reports1 reporting sourcesUpdated 2 h ago

Start here

AI digest

TrustGraph released an open-source policy-as-code tool that lets teams write a single policy file specifying who may read anything derived from each source, which agents may use it, and where each action sends output. It traces what each output was built from through every step and returns allow, needs approval or deny with a reason, using a deterministic decision path with no model involved. The project ships a Rust CLI, a gateway that sits between agents and MCP servers while holding their credentials, a local HTTP and MCP service, and Python integrations for LangChain/LangGraph, the OpenAI Agents SDK and the Claude Agent SDK. TrustGraph says the policy CLI and trustgraph serve have been through security review, while the gateway and Python integrations are new: two review rounds found and fixed real holes and a third is in progress, so they should not yet be put in front of real data. The tool requires Rust 1.90 or newer.

Generated from reports · updated 1 h ago

Report timeline

Follow the reports to see the event from different sides.

November 16
  1. Hacker News · AI
    TrustGraph launches policy-as-code for controlling where AI agent output can go

    TrustGraph is a new open-source tool that lets teams write one policy file specifying who may read anything derived from each source, which agents may use it, and where each action sends output. It records what each output was built from through every step and returns allow, needs approval or deny with a reason, using a deterministic decision path with no model involved. The project ships a Rust CLI, a gateway that sits between agents and MCP servers while holding their credentials, a local HTTP and MCP service, and Python integrations for LangChain/LangGraph, the OpenAI Agents SDK and the Claude Agent SDK; the gateway and Python integrations are described as new and not yet ready for real data.

Heat for this event

Not enough continuous observations yet to draw a trend.