The Agent Skills Directory

[SAFE]: The skill contains legitimate instructions for code exploration and does not include any malicious commands or unauthorized access attempts.\n- [PROMPT_INJECTION]: No prompt injection patterns, safety bypasses, or instruction overrides were detected. The skill includes a step to refine user questions into search queries, which reduces the risk of direct command injection.\n- [EXTERNAL_DOWNLOADS]: The skill does not initiate any external network requests or attempt to download third-party scripts or packages.\n- [DATA_EXFILTRATION]: There are no patterns suggesting data exfiltration; the skill operates on the local codebase and does not instruct the agent to send information to external domains.

explore