The Agent Skills Directory

[PROMPT_INJECTION]: The skill ingests untrusted content from developer forums, which creates a vulnerability to indirect prompt injection.
Ingestion points: The script scripts/community_signals.py fetches comments and posts from Hacker News and Reddit.
Boundary markers: No specific delimiters or safety instructions are used to separate untrusted data from the agent's instructions.
Capability inventory: The agent has access to Bash, Write, and WebSearch tools.
Sanitization: Content is cleaned of HTML tags but not of potentially malicious prompt instructions.
[SAFE]: The skill performs legitimate network operations to well-known services and vendor infrastructure.
It fetches public forum data from the Hacker News Algolia API.
It utilizes the vendor's proxy at app.gooseworks.ai to perform Reddit scraping via Apify.

community-signals