The Agent Skills Directory

[NO_CODE]: The skill is composed entirely of markdown instructions (SKILL.md) and does not contain any executable scripts, Python packages, Node.js modules, or binary files.\n- [SAFE]: No malicious patterns or security vulnerabilities were identified. The agent's capabilities are restricted to searching and reading code for analysis purposes. The instructions explicitly state that the agent is 'read-only' and must not make any code changes, providing a clear boundary for its operations. Potential risks associated with processing untrusted data (indirect prompt injection) are mitigated by the lack of write or network capabilities.

code-search