The Agent Skills Directory

[PROMPT_INJECTION]: The skill contains an indirect prompt injection surface as it guides the agent to process external configuration data.
Ingestion points: The skill uses Grep and Read tools to ingest data from various YAML files in the config/ directory (e.g., config/customize.yaml), as outlined in the 'Agentic Optimizations' section of SKILL.md.
Boundary markers: The instructions do not define delimiters or provide warnings to ignore embedded instructions within the processed data.
Capability inventory: The skill is granted Edit, Write, and TodoWrite tools, which allow for file system modifications across the configuration directory.
Sanitization: No sanitization or validation mechanisms are described for handling the ingested configuration content.

ha-entities