attenu-guard is an Apache-2.0 Python package that enforces least-privilege authorization when AI agents delegate work to sub-agents. It constrains child permissions to those of the parent and records audit logs that can be verified offline, with integrations for agent frameworks and MCP-based workflows.
attenu-guard sits in PulseGate's AI & LLM security category. It focuses on preventing delegated AI agents from exceeding their parent agent's permissions while preserving verifiable audit records. attenu-guard is an open-source project aimed at AI agent developers and security engineers. attenu-guard is open source under the Apache-2.0 license. It ships for the web, the command line, and API, and it can be self-hosted.
It is developed by attenu-io, and it first shipped in 2026. The project is developed in the open on GitHub with 44 commits in the last 90 days. Among its 7 catalogued features are permission inheritance, delegation controls, and least-privilege enforcement. It exposes integrations via an MCP server.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do