The Agent Skills Directory

[PROMPT_INJECTION]: The skill's archival instructions create an indirect prompt injection vulnerability surface. * Ingestion points: Data is ingested from files in the .ai_state/ directory. * Boundary markers: The instructions lack requirements for delimiters when appending data to conventions.md. * Capability inventory: The agent is authorized to move, write, and delete files. * Sanitization: There is no instruction to sanitize or validate content before it is processed.
[NO_CODE]: The skill is composed entirely of natural language instructions and does not provide any code or scripts.

smart-archive