predictit
Pass
Audited by Gen Agent Trust Hub on Feb 20, 2026
Risk Level: SAFE
Full Analysis
- Prompt Injection (SAFE): No instructions were found that attempt to override agent instructions or bypass safety filters.
- Data Exposure & Exfiltration (SAFE): The code does not access sensitive local files (like SSH keys or AWS credentials) or hardcoded secrets. It only fetches market data via an internal feed module.
- Remote Code Execution (SAFE): There are no patterns for downloading and executing remote scripts or packages.
- Indirect Prompt Injection (LOW): The skill processes untrusted data from an external API (PredictIt), which is a common surface for indirect injection.
- Ingestion points: External market questions (
market.question) and outcome names (o.name) are retrieved and formatted inindex.ts. - Boundary markers: IDs are wrapped in backticks (e.g.,
market.id), though market titles are displayed as bold text without rigorous escaping. - Capability inventory: The skill's logic is restricted to string formatting and data display; no file-writing, network-sending (beyond the feed), or command-execution capabilities are present in the analyzed files.
- Sanitization: No explicit sanitization of the API-provided strings is implemented in the display logic.
Audit Metadata