The Agent Skills Directory

[SAFE]: The skill's primary function is to enhance the agent's context by injecting documentation into project-specific files like CLAUDE.md or AGENTS.md. No malicious intent or unsafe operations were found.\n- [EXTERNAL_DOWNLOADS]: For Next.js projects, the skill utilizes 'npx @next/codemod@canary', which fetches documentation from Vercel's official repositories. This is considered a safe operational procedure as it targets a well-known and trusted technology company.\n- [COMMAND_EXECUTION]: The skill performs necessary command executions, such as running the 'npx' command for Next.js or a local Python script ('inject_fastapi_docs.py') for FastAPI. These commands are scoped to the project environment and are required for the skill's functionality.\n- [DATA_EXFILTRATION]: Analysis confirms that no sensitive data is accessed or transmitted to external servers. Data interaction is limited to reading project configuration for framework detection and writing documentation to local files.\n- [PROMPT_INJECTION]: The instructions in the skill body guide the agent on framework detection and execution steps. These are benign operational instructions and do not attempt to bypass safety filters or override core agent behaviors.

inject-docs