deep-research

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE] (SAFE): No malicious patterns, obfuscation, or data exfiltration attempts were detected.
  • Credential Security: The skill correctly implements API key retrieval via environment variables, avoiding the risk of hardcoded secrets.
  • Network Operations: The script exclusively communicates with official Google API endpoints (ai.google.dev), which are categorized as trusted.
  • File Handling: Local file system operations are limited to reading user-specified research context and writing output reports, matching the tool's intended purpose.
  • Indirect Prompt Injection (LOW): The skill features a data ingestion surface where external file content is processed by the AI. 1. Ingestion points: scripts/research.py via the --file parameter. 2. Boundary markers: Absent in the prompt construction. 3. Capability inventory: Recursive file reading, network uploads to Gemini APIs, and local file creation. 4. Sanitization: None. This risk is acknowledged but considered inherent to the skill's primary function of document-based research.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:03 PM