{"id":"agent-governance-toolkit","name":"Agent Governance Toolkit","maker":"Microsoft","url":"https://github.com/microsoft/agent-governance-toolkit","repo":"microsoft/agent-governance-toolkit","docs":"https://microsoft.github.io/agent-governance-toolkit","package":"pypi:agent-governance-toolkit","summary":"A Microsoft toolkit that checks every tool call, message and delegation of an AI agent against policy before it runs.","details":"Wraps agent tools so each call is evaluated against a YAML policy, can be denied or sent for human approval, and is written to an audit trail; it also covers agent identity and sandboxing. It ships as Python, npm and NuGet packages with adapters for agent frameworks and a Claude Code plugin. The code is open source under the MIT licence.","open_source":true,"hosted":false,"also":["integrations"],"verified":"2026-11-15","category":"safety","category_name":"Guardrails and security","page":"https://modelmoot.com/tools/agent-governance-toolkit","reviewed":"2026-11-15","checked":{"at":"2026-11-15T13:03:28.711Z","address_answers":true,"repository":{"at":"2026-11-15T13:03:28.711Z","archived":false,"licence":"MIT","language":"Python","last_change":"2026-11-14T23:01:59.000Z","latest_release":{"tag":"v4.1.0","published":"2026-06-09T23:11:52.000Z"}}},"note":"Descriptions are the site's own. Facts under \"checked\" were read automatically from the tool's address and public repository."}