The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests content from project files, which serves as an attack surface for instructions embedded in external data.
Ingestion points: The skill reads about-me.md and voice.md to inform its content generation and pillar suggestions.
Boundary markers: No specific delimiters or "ignore instructions" directives are provided to separate the external file content from the agent's logic.
Capability inventory: The skill has the capability to write files to the filesystem (e.g., content-matrix-YYYY-MM-DD.md).
Sanitization: There are no instructions to sanitize or validate the contents of the identity files before they are processed.
[PROMPT_INJECTION]: The skill uses directive language designed to override default agent behaviors such as summarization.
Evidence: The instruction CRITICAL: Auto-start on load combined with Do not summarise uses high-priority markers to force specific execution behavior upon triggering.

content-matrix