{"id":"snyk-agent-scan","name":"Snyk Agent Scan","maker":"Snyk (originally Invariant Labs)","url":"https://github.com/snyk/agent-scan","repo":"snyk/agent-scan","package":"pypi:snyk-agent-scan","summary":"A command-line scanner that checks agents, MCP servers and agent skills for security problems such as tool poisoning.","details":"Finds the agent clients, MCP servers and skills installed on a machine, reads the tool descriptions and skill files they expose, and flags issues such as prompt injection, sensitive data handling and hidden malware. To read tools it may start configured MCP servers. It was published earlier as mcp-scan by Invariant Labs. The code is open source under Apache 2.0.","open_source":true,"hosted":false,"speaks":["mcp","agent-skills"],"also":["mcp-tooling"],"verified":"2026-11-15","category":"safety","category_name":"Guardrails and security","page":"https://modelmoot.com/tools/snyk-agent-scan","reviewed":"2026-11-15","checked":{"at":"2026-11-15T16:10:15.342Z","address_answers":true},"note":"Descriptions are the site's own. Facts under \"checked\" were read automatically from the tool's address and public repository."}