The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from local project files to drive its execution logic, which creates a surface for indirect prompt injection.
Ingestion points: The skill reads from plan.md, spec.md, and tech-stack.md to identify tasks, requirements, and dependencies (documented in SKILL.md and resources/implementation-playbook.md).
Boundary markers: No explicit delimiters or instructions are provided to the agent to distinguish between task data and potential malicious instructions within those files.
Capability inventory: The skill instructs the agent to execute shell commands for testing and linting (pytest, ruff, mypy) and perform version control operations (git commit, git notes) as detailed in resources/implementation-playbook.md.
Sanitization: No sanitization or validation of the input file content is implemented before processing.

workflow-patterns