polymarket-user-analyzer
Pass
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The script
scripts/analyze_user.jsperforms network requests topolymarket.comanddata-api.polymarket.com. These are official, well-known domains for the service being analyzed, and the operation is used to fetch public profile and trading history data. - [COMMAND_EXECUTION]: The skill uses Node.js built-in modules to perform operations. Specifically,
scripts/analyze_user.jsusesfetchfor network communication andfs.writeFileSyncto save analysis reports to the local file system. These actions are transparently documented and consistent with the skill's primary purpose. - [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection due to its ingestion of untrusted external content.
- Ingestion points: User profile HTML is scraped from
polymarket.comand trading activity JSON is fetched fromdata-api.polymarket.cominsidescripts/analyze_user.js. - Boundary markers: There are no boundary markers or instructions to ignore embedded content when processing or displaying retrieved data.
- Capability inventory: The skill has the capability to write to the local file system (
fs.writeFileSync) and log information to the terminal, which could be leveraged if the output is later processed by an agent. - Sanitization: External strings, including trading market titles and usernames, are interpolated directly into console logs and report files without escaping or validation.
Audit Metadata