The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes project specifications and requirement documents to generate plans. This is an ingestion point for untrusted data that could potentially contain instructions to influence the task breakdown process.
Ingestion points: Reads project specifications and requirements during the planning phase.
Boundary markers: None explicitly defined in the prompt templates.
Capability inventory: Skill invocation of attune:project-execution and file writing to docs/implementation-plan.md.
Sanitization: Standard LLM processing without specific sanitization routines.
[COMMAND_EXECUTION]: The skill contains logic for an 'Automatic Trigger' that invokes the next workflow phase (attune:project-execution) without user confirmation. While this increases agent autonomy, it is constrained to the vendor's own ecosystem and described as a lightweight workflow checkpoint.

project-planning