The Agent Skills Directory

[SAFE]: The skill provides behavioral instructions for the agent to minimize code bloat. It explicitly reinforces security best practices, stating that input validation at trust boundaries and error handling are non-negotiable. No malicious patterns or obfuscation techniques were detected in the instructions or metadata.\n- [NO_CODE]: The skill contains only documentation and platform configuration, with no embedded scripts or binary executables.\n- [INDIRECT_PROMPT_INJECTION]: The skill interacts with external codebase data. Evidence chain: 1. Ingestion points: Codebase files accessed via Read, Grep, and Glob tools. 2. Boundary markers: Absent. 3. Capability inventory: Limited to read-only operations (Read, Grep, Glob) as per the allowed-tools configuration. 4. Sanitization: Absent. The risk is considered negligible due to the restricted toolset and explicit security guardrails.

yagni-gate