codebase-audit

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it processes untrusted documentation and codebase files to build its analysis.
  • Ingestion points: Reads project-level documentation such as README.md, AGENTS.md, and source code files during the architectural investigation and audit phases.
  • Boundary markers: The instructions do not define delimiters or specific 'ignore' commands to isolate external content from the agent's core instructions.
  • Capability inventory: Uses code navigation and file reading tools to search and retrieve content from the repository.
  • Sanitization: No validation or sanitization is performed on the content of the ingested files.
  • [DATA_EXFILTRATION]: The skill explicitly directs the agent to identify and report sensitive information, including API keys, account IDs, and environment-specific hostnames. While this is the intended purpose of the audit, it facilitates the exposure of credentials within the project context.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 03:36 PM
Security Audit — agent-trust-hub — codebase-audit