analyze-project
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from previous session artifacts which could technically contain malicious instructions if a prior session was compromised.
- Ingestion points: Markdown and JSON files within the
~/.gemini/antigravity/brain/directory. - Boundary markers: The skill contains a dedicated 'Measurement and privacy boundaries' section that explicitly instructs the agent to ignore commands embedded in archived conversations.
- Capability inventory: The skill is limited to reading files, listing directory contents, and writing a markdown report. It lacks network access, shell execution, or privilege escalation capabilities.
- Sanitization: The instructions mandate redacting private excerpts and credentials during the analysis process.
- [DATA_EXPOSURE]: The skill is designed to read session logs which may contain sensitive project information.
- The skill includes strong defensive instructions to redact credentials and keep all analysis output local, which mitigates the risk of sensitive data exposure during the reporting phase.
Audit Metadata