brand-strategy
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill interacts with user-supplied data and external research, creating an entry point for potential indirect prompt injection attacks.
- Ingestion points: The skill reads user-provided context, brand tracking data, and external research as specified in SKILL.md and references/skill-spec.json.
- Boundary markers: The references/output-contract.md establishes clear sections, and instructions require explicit separation of facts from assumptions.
- Capability inventory: The skill can execute internal scripts (dynamic_router.py, neural_router.py) and perform network research.
- Sanitization: Strict evidence policies in references/skill-spec.json mandate the classification of claims and prohibit the fabrication of data or tool outputs.
- [COMMAND_EXECUTION]: The skill instructions and specification refer to local scripts for advanced routing functionality.
- Evidence: SKILL.md and references/skill-spec.json direct the agent to use ../../scripts/dynamic_router.py and ../../scripts/neural_router.py for handling cross-functional dependencies and agent selection.
Audit Metadata