The Agent Skills Directory

[SAFE]: Comprehensive analysis of the provided documentation reveals no security risks. The skill serves as a knowledge base for developers working on program analysis tools.
[COMMAND_EXECUTION]: The skill mentions standard development commands like node scripts/build.js and tsc clean within a validation checklist. These are documented as manual steps for developers to verify their work and do not involve the execution of untrusted or remote code by the AI agent.
[PROMPT_INJECTION]: No evidence of instructions designed to bypass safety filters or override agent behavior. The language is technical and focused on software engineering concepts such as reachability, data dependencies, and interprocedural summaries.

gitnexus-taint-analysis