production-excellence
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill serves as an orchestration point that ingests and processes evidence data from various specialist domains (readiness, migration, resilience, etc.) to produce launch decisions. This ingestion of data from external specialist skills and user-provided launch details creates a standard attack surface for indirect prompt injection, where malicious instructions embedded in the evidence could attempt to influence the evaluation logic.\n
- Ingestion points: The skill reads domain-specific evidence as specified in
references/evidence-packet.md.\n - Boundary markers: Data is structured within Markdown tables and templates, providing organizational separation between different evidence domains.\n
- Capability inventory: The skill's primary capabilities are logical evaluation and the generation of text-based records like the
handoff-record.md. It does not execute shell commands or perform network operations.\n - Sanitization: No specific filtering or escaping is applied to the gathered evidence content, relying on the structured nature of the report templates to manage data flow.
Audit Metadata