The Agent Skills Directory

[SAFE]: The skill is designed for project maintenance and organization. Its operations are restricted to searching local files within the project repository to identify functional overlap between skills and agents.
[PROMPT_INJECTION]: The skill ingests data from external files such as SKILL.md and .claude/agents/*.md. This presents a potential surface for indirect prompt injection if those files contain malicious instructions. However, because the skill is intended to read these files for auditing purposes and the scope is limited to the local development environment, the risk is minimal.
[DATA_EXFILTRATION]: No network operations or data exfiltration patterns were detected. The skill does not use tools like curl or wget and does not attempt to send data to external domains.

skill-preflight