The Agent Skills Directory

[SAFE]: The skill facilitates an autonomous development workflow using standard tools like Bash and file system access. This behavior is consistent with its intended purpose. While it processes repository data that could theoretically contain indirect prompt injections, this is an inherent risk of the task category and no specific exploits were found. \n- Ingestion points: Repository files read during the inspection phase in SKILL.md. \n- Boundary markers: Not explicitly defined in the instructions. \n- Capability inventory: Read, Write, Bash, Grep, and Glob tools. \n- Sanitization: Not explicitly defined in the instructions.

autopilot