The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were detected. The skill implements read-only data retrieval from official Polymarket API endpoints (gamma-api.polymarket.com and clob.polymarket.com).
[PROMPT_INJECTION]: The skill identifies a potential indirect prompt injection surface by processing user-generated market descriptions from Polymarket. However, this is addressed through developer warnings and sanitization.
Ingestion points: Market metadata fetched in scripts/scan_markets.py from the Gamma API.
Boundary markers: SKILL.md contains an explicit warning to treat market names as untrusted data.
Capability inventory: The scripts are limited to data fetching and printing JSON to standard output. They lack dangerous capabilities like file writing or subprocess execution based on the fetched data.
Sanitization: scripts/scan_markets.py includes a sanitize_text function that strips control characters and limits the length of external content.

polymarket-scanner