{"id":"invariant","name":"Invariant Guardrails","maker":"Invariant Labs (part of Snyk)","url":"https://invariantlabs.ai","repo":"invariantlabs-ai/invariant","docs":"https://invariantlabs-ai.github.io/docs/mcp-scan/guardrails-reference/","summary":"A rule-based guardrailing layer that checks agent traces and tool calls against policies written in a Python-like language.","details":"Policies describe forbidden patterns of messages and tool calls, such as data flowing from a private source to an external send, and are enforced by the Invariant Gateway, which sits as a proxy in front of an LLM provider or MCP servers. The code is open source under Apache 2.0. The Invariant Labs site states that Snyk has acquired the company.","open_source":true,"hosted":false,"speaks":["mcp"],"verified":"2026-11-15","category":"safety","category_name":"Guardrails and security","page":"https://modelmoot.com/tools/invariant","reviewed":"2026-11-15","checked":{"at":"2026-11-15T14:03:44.997Z","address_answers":true,"repository":{"at":"2026-11-16T00:27:04.284Z","archived":false,"licence":"Apache-2.0","language":"Python","last_change":"2026-01-12T12:50:28.000Z","latest_release":null}},"note":"Descriptions are the site's own. Facts under \"checked\" were read automatically from the tool's address and public repository."}