prediction-market-arbitrage-zh
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill makes network requests to
https://api.aisa.one/apis/v1inscripts/arbitrage_finder.pyandscripts/prediction_market_client.pyto fetch market data and prices. This is normal functionality for a data-driven skill. - [DATA_EXFILTRATION]: The skill uses the
AISA_API_KEYenvironment variable to authenticate requests sent to the vendor domainapi.aisa.one. No unauthorized access to local sensitive files or credentials was observed. - [INDIRECT_PROMPT_INJECTION]: The skill ingests external data from the AIsa API, which could potentially contain malicious instructions.
- Ingestion points: Market titles and metadata are retrieved via API calls in
scripts/arbitrage_finder.pyandscripts/prediction_market_client.py. - Boundary markers: None identified.
- Capability inventory: The skill uses
urllib.requestfor network retrieval but lacks capabilities for code execution (eval,exec) or file system modification. - Sanitization: Standard JSON decoding is used, with no additional sanitization for content intended for agent context.
Audit Metadata