flow-qa-orchestrator
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from PRDs and technical specifications which creates a surface for indirect prompt injection.
- Ingestion points: The agent reads document files (e.g.,
./docs/PRD.md) or text provided by the user in 'Fase 1'. - Boundary markers: There are no instructions for the agent to use delimiters or ignore instructions embedded within the provided documents.
- Capability inventory: The agent has the ability to create directories, write files (
qa_test_plan.md,qa_session.json), and invoke subagents (flow-qa-task-runner). - Sanitization: The skill does not instruct the agent to sanitize or filter content extracted from the requirement documents before processing.
Audit Metadata