consult-claude
Pass
Audited by Gen Agent Trust Hub on Aug 16, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process outputs from an external source (Claude Code), which introduces a potential surface for indirect prompt injection if the collaborator's output contains malicious instructions.
- Ingestion points: The agent processes Claude's responses and follow-up evidence in Step 2 and Step 3.
- Boundary markers: The instructions include explicit safeguards, requiring that the agent 'Cross-reference suggestions with the repository before acting' and 'Do not apply a suggestion until it has been checked against the actual codebase.'
- Capability inventory: The skill uses the
$claude-printtool to send questions andupdate_planto manage workflow state. - Sanitization: The skill relies on the agent's verification against the actual repository state to mitigate incorrect or malicious suggestions.
Audit Metadata