The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill is distributed via the NPM registry and installed using the npx command.
[COMMAND_EXECUTION]: The skill utilizes the Bash tool to perform essential research and validation tasks:
Executes gh repo view to collect repository metadata such as star counts and contributor activity.
Uses curl to verify live product endpoints and landing page content.
Conditionally runs codex exec in read-only mode to perform secondary analysis on gathered data.
[PROMPT_INJECTION]: The skill implements an indirect prompt injection surface by ingesting untrusted data from the web to generate its audit reports.
Ingestion points: External landing pages and stats APIs (Phase 0), and search engine results from WebSearch and WebFetch (Phase 2).
Boundary markers: No explicit instructions for delimiters or escaping are provided for the content fetched from the web.
Capability inventory: The agent possesses the Bash tool for shell execution, the Write tool for file creation, and the Agent tool for spawning subagents.
Sanitization: While the skill lacks technical sanitization (escaping) for fetched data, it incorporates a mandatory 'Adversarial review loop' (Phase 6) using subagents to validate the consistency and feasibility of the generated document.

validate-market