The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it ingests untrusted data from external websites and review platforms.
Ingestion points: Data enters the agent context via JSON files generated by fetch_pages.py and render_page.mjs (e.g., ${WORKSPACE}/competitor_pages.json, ${WORKSPACE}/g2_reviews.json).
Boundary markers: Absent. The instructions do not specify delimiters (like XML tags or triple quotes) or provide warnings to the agent to ignore instructions embedded within the scraped content.
Capability inventory: The skill uses python3, node, npx (for Chromium installation), and curl (as a fallback mention for Apify) to interact with the system and network.
Sanitization: Absent. The scripts extract raw text, headings, and meta-descriptions without filtering for potentially malicious instructions.
[EXTERNAL_DOWNLOADS]: The skill downloads the Chromium browser using npx playwright install chromium. While this is a well-known tool from a reputable source, it involves downloading and installing an executable binary at runtime.
[COMMAND_EXECUTION]: The skill relies on executing local Python and Node.js scripts to perform its primary functions. It also provides a command for manual installation of browser dependencies.

battlecard-generator