{"id":"mindgard","name":"Mindgard","maker":"Mindgard","url":"https://mindgard.ai","package":"pypi:mindgard","summary":"A commercial platform that tests AI systems and agents by attacking them as an adversary would.","details":"Maps AI systems, agents and tools, runs adaptive attacks to find exploitable weaknesses, suggests and retests fixes, and adds runtime threat response. A command-line client is published on PyPI under the MIT licence; the platform needs an account and is not open source.","open_source":false,"hosted":true,"verified":"2026-11-15","category":"safety","category_name":"Guardrails and security","page":"https://modelmoot.com/tools/mindgard","reviewed":"2026-11-15","checked":{"at":"2026-11-15T15:07:43.409Z","address_answers":true},"note":"Descriptions are the site's own. Facts under \"checked\" were read automatically from the tool's address and public repository."}