The Agent Skills Directory

[SAFE]: The skill consists of Markdown instructions and YAML configuration for prompt engineering. It does not contain any executable scripts, binary files, or commands that interact with the host system's shell or network.
[INDIRECT_PROMPT_INJECTION]: The skill processes external files (SOURCE_BIBLE.md and PRODUCT_BRIEF.md). While these represent an ingestion surface for untrusted data, the skill has no dangerous capabilities (like network access or high-privilege shell execution) to exploit, limiting risk to output manipulation. 1. Ingestion points: SOURCE_BIBLE.md and PRODUCT_BRIEF.md (SKILL.md). 2. Boundary markers: None specified in instructions. 3. Capability inventory: No restricted tools, subprocess calls, or network operations are requested. 4. Sanitization: None observed.

game-concept