Observed arrival · 2026-09-06
Linmas, a Security Review Loop for AI Coding Agents
An open-source collection of defensive security skills that helps AI coding agents review code, explain risks, and suggest concrete fixes before deployment.
Field notes
Linmas structures security review around a named target—such as a diff, endpoint, feature, or deployment plan—and routes that target to a focused skill rather than one broad security prompt. Its example output pairs severity and impact with a fix and a verification step, including parameterized queries and regression tests for SQL injection. The homepage lists 11 skills, 7 native MCP tools, 3 policy packs, an Apache-2.0 license, and a current v0.8.0 release.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue