2026-legal-research-agent
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [Indirect Prompt Injection] (LOW): The script
validate-scraped-data.tsingests untrusted data from scraped JSON files.\n - Ingestion points: Files in
src/data/scraped/states/are read viafs.readFileSyncand parsed as JSON in thevalidateStateDatafunction.\n - Boundary markers: There are no delimiters or 'ignore embedded instructions' warnings present when the script reads and processes the 'content' field.\n
- Capability inventory: The script's capabilities are limited to reading local files (
fs.readFileSync) and printing results to the console. No network requests, shell command execution, or file write operations were detected.\n - Sanitization: No escaping, filtering, or sanitization of external text content is performed beyond checking for JSON structure and basic string patterns like '404'.\n- [Unverifiable Dependencies & Remote Code Execution] (SAFE): The script uses
npx tsxin its shebang. Whilenpxfetches packages from the npm registry,tsxis a standard development tool for executing TypeScript, and no unusual or untrusted packages are explicitly listed for installation.
Audit Metadata