AgentTrust is a verifiable zero-trust control plane, cryptographic black box and policy proxy that sits between autonomous LLM agent runtimes and the external tools they call. It intercepts every proposed action, screens parameters for prompt injection and data leaks, checks real-time blast radius, enforces fine-grained capability policies, and gates dangerous actions behind approval. The project targets EU AI Act Article 12 and 14 compliance logging and is dual licensed MIT and commercial.
Be the first to rate AgentTrust. Your feedback helps other users decide.
Route agent tool calls through the proxy so nothing reaches a shell, database or card without a policy check.
Let it inspect arguments for injected instructions and outbound data leaks before the call runs.
Use the cryptographic black box and compliance logging to answer what the agent did and why it was allowed.
Similar tools in the same category — compare side-by-side
unspin shows you what your coding agent's shell commands are doing instead of a spinner. When Claude...
Decido returns calibrated yes/no, pick-one and rating decisions about a document on a laptop CPU. Yo...
LexFrame is a Python toolkit for interpreting language-model residual streams in their native semant...
Guides and comparisons about code tools
Nine launches that draw the line around an agent: policy proxies, guard hooks, semantic permissions, sandboxes, worktrees and a proof checker.
Self-hosted agent platforms, permission kernels, incident agents and simulated environments for teams that need agents inside their own infrastructure.
Ten Hacker News launches that treat agent context, persistent memory and token accounting as first-class problems instead of afterthoughts.
Eleven recent open-source and self-hosted projects for building, sandboxing and deploying coding agents you own outright rather than rent.
Eleven recent projects that turn the Model Context Protocol from a neat demo into deployable infrastructure: registries, compiled servers, and gateways.
