The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and refactor existing skill files which may contain untrusted content from third-party sources.
Ingestion points: The skill uses the Read tool to analyze existing SKILL.md files during the Assessment and Identification phases.
Boundary markers: The instructions lack explicit delimiters or "ignore embedded instructions" warnings when processing the content of the skills being optimized.
Capability inventory: The skill utilizes Write, Edit, and Glob tools across its scripts to restructure directories and refactor file content.
Sanitization: There is no prescribed sanitization or validation of the input content before the refactoring and writing operations occur.
[COMMAND_EXECUTION]: The optimization methodology encourages extracting mechanical operations into local scripts (e.g., scripts/validate.py, scripts/upload.py) and provides specific patterns for the agent to execute these scripts via the bash tool. This assumes the extracted or generated scripts are secure and do not contain malicious payloads.

skill-optimizer