adr-discovery

Pass

Audited by Gen Agent Trust Hub on May 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a strict 'zero hallucination' policy, requiring human confirmation for every fact before it is written to the project's documentation. This reduces the risk of accidental prompt-driven misinformation or data corruption.
  • [COMMAND_EXECUTION]: The skill utilizes legitimate shell commands such as git log and npx likec4 to analyze project history and start a local architecture visualization server. These operations are standard for development workflows and targeted at a local environment.
  • [EXTERNAL_DOWNLOADS]: The skill recommends the installation of likec4 via package managers (npm) if it is missing from the development environment. LikeC4 is a well-known, industry-standard tool for architecture-as-code.
  • [DATA_EXFILTRATION]: No unauthorized network operations or data exfiltration attempts were identified. Network activity is confined to the LikeC4 development server running on a local port (default 5173) for real-time diagram rendering.
Audit Metadata
Risk Level
SAFE
Analyzed
May 10, 2026, 10:00 PM
Security Audit — agent-trust-hub — adr-discovery