The Agent Skills Directory

[SAFE]: The skill contains instructional content for AI product engineering and does not perform any automated actions, downloads, or sensitive file access.
[INDIRECT_PROMPT_INJECTION]: The skill includes a surface for indirect injection as it processes untrusted data from model calls and retrieval (Ingestion: model/tool calls in SKILL.md). While explicit boundary markers are not defined in the instructions, the skill environment includes 'workspace-write' capabilities. The instructions specifically mandate sanitization through 'schema-safe parsing' and 'structured output validation' to mitigate these risks.

ai-engineer