The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection by processing untrusted data from project files.
Ingestion points: The workflow reads external files including spec.md, conductor/tracks.md, conductor/product.md, and conductor/tech-stack.md (Step 1).
Boundary markers: Instructions lack explicit delimiters or safety warnings to ignore embedded instructions within the source files.
Capability inventory: The skill possesses read_file and write_file capabilities, which are used to transform input data into a new plan.md file.
Sanitization: There is no evidence of sanitization, filtering, or validation of the content retrieved from input files before it is used to generate the plan and metadata updates.

loop-planner