used-car-market-comparison

Pass

Audited by Gen Agent Trust Hub on Oct 20, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests vehicle listing data from third-party sources (CarMax, Autotrader, Cars.com) via an API for comparison and analysis. This external content could theoretically contain instructions intended to influence the agent's behavior during the comparison process.
  • Ingestion points: External data is retrieved through the scripts/crawlora.sh helper script from api.crawlora.net.
  • Boundary markers: The instructions do not define specific delimiters to isolate the external listing data from the agent's analytical instructions.
  • Capability inventory: The skill is capable of executing network requests via curl through the provided script, but lacks file-system write or persistent execution capabilities.
  • Sanitization: The helper script performs basic path validation and uses curl's parameter encoding to mitigate command injection risks.
  • [COMMAND_EXECUTION]: The skill uses a local bash script (scripts/crawlora.sh) to perform API calls. The script incorporates security-focused design elements:
  • Credential Protection: It uses mktemp and a curl config file to provide the API key via a header, preventing the secret from appearing in process listings or command history.
  • Input Validation: It enforces strict format checks on the CRAWLORA_API_KEY and uses a whitelist of allowed API routes via regex matching to prevent unauthorized path traversal or usage of unrelated API endpoints.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 20, 2026, 06:45 AM