product-research

Pass

Audited by Gen Agent Trust Hub on Jul 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were detected in the skill instructions or supporting scripts.
  • [SAFE]: The Python scripts (insight_confidence_scorer.py, method_recommender.py, and screener_validator.py) use only Python standard libraries. They do not perform network requests, file system writes, or execution of arbitrary code.
  • [SAFE]: Indirect Prompt Injection Surface Analysis (Category 8):
  • Ingestion points: Research questions, evidence claims, and screener items are ingested via JSON input files in scripts/method_recommender.py, scripts/insight_confidence_scorer.py, and scripts/screener_validator.py.
  • Boundary markers: Input data is interpolated directly into text outputs without explicit delimiters.
  • Capability inventory: The skill's capabilities are limited to local data transformation and rendering; there are no network, file-write, or shell-execution capabilities associated with the processing of untrusted data.
  • Sanitization: Input strings are not sanitized before rendering, but the risk is mitigated by the restricted environment and lack of dangerous tools.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 24, 2026, 01:16 PM
Security Audit — agent-trust-hub — product-research