The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data and follows external links.
Ingestion points: Reads local PRD files (e.g., docs/PRD.md), GitHub issues via the gh CLI, and content from external URLs linked within documents (documented in SKILL.md under Workflow step 3).
Boundary markers: No explicit delimiters or instructions are used to distinguish between data to be processed and instructions to be ignored within the ingested content.
Capability inventory: The skill can read and write local files (README updates), execute shell commands (gh, bun test, pytest), and create GitHub issues.
Sanitization: The skill lacks validation or sanitization of content retrieved from external links or PRD files before using it to guide the agent's actions.
[COMMAND_EXECUTION]: The skill utilizes local command-line tools to perform its auditing and verification tasks.
Executes gh issue list to audit the status of user stories and milestones.
Executes bun test or pytest to verify implementation quality.
While these commands are central to the skill's purpose, they provide a surface for executing code on the host environment based on configuration found in the repository.

prd-completion-reviewer