skills/ben2pc/auriga-cli/deep-review/Gen Agent Trust Hub

deep-review

Pass

Audited by Gen Agent Trust Hub on Jul 2, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill enforces a 'Mandatory New Context Isolation' policy for every sub-agent, which effectively mitigates the risk of indirect prompt injection by ensuring that potentially malicious instructions in a pull request cannot influence the agent's broader logic or persistent state.
  • [EXTERNAL_DOWNLOADS]: Documentation is fetched exclusively from trusted domains such as Anthropic and OpenAI to ensure compliance with the latest plugin and skill standards.
  • [COMMAND_EXECUTION]: System commands are limited to legitimate project utilities like git and the GitHub CLI for processing pull requests, as well as a local python-based metadata validator.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 2, 2026, 03:03 AM
Security Audit — agent-trust-hub — deep-review