research-market
Pass
Audited by Gen Agent Trust Hub on Jun 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized data access detected. The skill implements robust orchestration and safety principles.
- [DATA_EXFILTRATION]: No network operations or sensitive data access patterns were identified. The skill explicitly prohibits agents from including credentials, API keys, tokens, or PII in context, prompts, or logs (Principle 4).
- [PROMPT_INJECTION]: The instructions are directive for agent coordination and do not attempt to bypass safety filters or override system-level constraints. Role-playing is used for task-specific personas without adversarial intent.
- [INDIRECT_PROMPT_INJECTION]: The skill processes project-specific data to inform research.
- Ingestion points: Reads dependency manifests (e.g., package.json), roadmap files, research artifacts, and user feedback files (inferred from role definitions).
- Boundary markers: Explicit markers for untrusted data are absent.
- Capability inventory: Agent spawning (Agent tool), inter-agent messaging (SendMessage), and file management (docs/progress/ and docs/research/).
- Sanitization: Not explicitly defined, though agents are instructed to distinguish facts from inferences and label confidence levels.
Audit Metadata