Defensive Security Reviewer
Reviews source code defensively to uncover trust-boundary vulnerabilities and implementation risks without unauthorized execution.
The Defensive Security Reviewer skill provides structured, read-only security analysis of codebases within Claude Code. By enforcing strict operational boundaries that restrict execution to read and search functions, this skill systematically assesses input validation, context requirements, trust boundaries, and implementation risks. It helps engineering teams identify potential security vulnerabilities, adherence to project security conventions, and architectural weaknesses early in the development lifecycle while strictly avoiding intrusive operations like exploitation or unauthorized deployments.
主な機能
010 GitHub stars
02Defensive, read-and-search-only inspection of target source files
03Trust-boundary and privilege-boundary risk assessment
04Systematic verification workflow preventing scope creep
05Structured gap and blocker reporting when security context is incomplete
06Strict operational denial of destructive exploits or deployments
ユースケース
01Auditing newly implemented features or pull requests for boundary validation flaws
02Checking existing legacy modules against internal security conventions and trust boundaries
03Conducting read-only vulnerability reviews prior to production deployment