The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands including gh api, gh repo view, and git remote to interact with GitHub repositories. It also provides a node -e command to execute a local script for image rendering.
[EXTERNAL_DOWNLOADS]: The instructions reference well-known external dependencies and services, specifically the puppeteer Node.js library and npx @vercel/og for image generation.
[DYNAMIC_EXECUTION]: The skill generates a specific Node.js script at runtime to handle the conversion of HTML templates into PNG images via a headless browser. This execution is confined to the local environment and the specific task of rendering.
[INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it ingests data from local repository files to populate templates.
Ingestion points: Project name and description are parsed from package.json, Cargo.toml, and pyproject.toml in Step 1 of the generation procedure.
Boundary markers: The skill does not explicitly define boundary markers or 'ignore' instructions for the data interpolated into the HTML template.
Capability inventory: The agent possesses the capability to execute shell commands (node -e, gh api) and perform file system operations.
Sanitization: There is no explicit mention of sanitizing or escaping the metadata retrieved from manifest files before it is rendered into the HTML output.

social-preview