request-optimizer
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill consists of comprehensive documentation and reference files that establish a logic-based routing system. No malicious code, hardcoded credentials, or unauthorized network operations were identified. The external services referenced (Notion, Supabase, Vercel, etc.) are well-known platforms.\n- [INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface by ingesting user-provided requests for analysis and routing through coordinated agents and tools. However, the architecture utilizes mandatory approval gates for high-risk actions to maintain human-in-the-loop control.\n
- Ingestion points: Processes user requests provided via the CLI interface and future session context (as described in SKILL.md and references/integration-interfaces.md).\n
- Boundary markers: Implements a structured recommendation format that explicitly asks for user approval before proceeding with execution paths.\n
- Capability inventory: Capable of orchestrating 27 agents, 27 skills, and 14 MCP servers for tasks ranging from file manipulation to cloud API interactions.\n
- Sanitization: Uses a 5-point specificity analysis framework (documented in references/analysis-framework.md) to clarify user intent and identify subtasks before any tools are invoked.
Audit Metadata