execute
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-supplied task descriptions and use them to orchestrate complex projects.
- Ingestion points: User-provided task descriptions and additional instructions provided at skill invocation (SKILL.md).
- Boundary markers: None explicitly defined to separate user input from orchestration instructions.
- Capability inventory: Mentions access to basic tools (file read/write, code search, command execution) and external MCPs (semantic search, browser automation via Playwright, and technical documentation).
- Sanitization: No specific sanitization or validation logic is defined for the incoming user instructions.
Audit Metadata