The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it processes external task lists and implementation plans.
Ingestion points: Reads implementation plans from the filesystem (e.g., docs/implementation-plan.md) and task descriptions to drive the execution loop (SKILL.md).
Boundary markers: Absent. The skill does not define specific delimiters or instructions for the agent to ignore potentially malicious instructions embedded within the ingested project plans.
Capability inventory: The skill performs file system modifications, executes test runners (pytest), and runs build automation tools (make) (SKILL.md).
Sanitization: Absent. No validation or escaping of external plan content is documented before processing.
[COMMAND_EXECUTION]: The framework requires the execution of shell commands for validation and testing purposes.
Evidence: The instructions direct the agent to run pytest -v, make lint, make typecheck, make test, and make coverage to verify implementation quality (SKILL.md).

project-execution