The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands including awk, sed, and npx to parse local configuration files and fetch information from remote sources. Specifically, it reads skill descriptions from the local configuration path ~/.config/everything-claude-code/skills and interacts with a vendor repository to facilitate the audit.
[DYNAMIC_EXECUTION]: Employs an inline Python script via uv run to process YAML data from upstream/distribution.yaml. The script uses yaml.safe_load to handle the data securely.
[EXTERNAL_DOWNLOADS]: Fetches skill listings from the vendor repository (ValorVie/ai-dev-skills) using the npx skills tool.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests and analyzes descriptions from external upstream skills. 1. Ingestion points: Skill descriptions and metadata are read from local paths in ~/.config/everything-claude-code/skills/ (SKILL.md). 2. Boundary markers: No explicit delimiters or isolation warnings are used when processing the ingested descriptions. 3. Capability inventory: The skill has the ability to execute shell commands and Python scripts. 4. Sanitization: Employs yaml.safe_load for structured data ingestion.

custom-skills-ecc-analyze