The Agent Skills Directory

[SAFE]: No malicious patterns, persistence mechanisms, or unauthorized activities were detected. The skill adheres to its defined purpose of providing prompt optimization advice.\n- [DATA_EXFILTRATION]: The skill reads project configuration files such as package.json, go.mod, and pyproject.toml to determine the technology stack. This information is used locally to provide relevant prompt suggestions and is not transmitted to external servers.\n- [COMMAND_EXECUTION]: Although the skill generates prompts containing ecosystem-specific commands (e.g., /plan, /tdd, /verify), it is explicitly prohibited from executing these commands itself. Its function is limited to text generation.\n- [PROMPT_INJECTION]: While the skill ingests untrusted user prompts for optimization, the risk of indirect prompt injection is mitigated by the skill's strict advisory-only constraint, which prevents the agent from executing any resulting malicious instructions.

prompt-optimizer