solid-generic
Pass
Audited by Gen Agent Trust Hub on Feb 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill mandates an orchestrational workflow using
TeamCreateto spawn specialized agents (fuse-ai-pilot:explore-codebase,fuse-ai-pilot:research-expert) and utilizes themcp__context7__query-docstool. These are standard architectural patterns for the 'fusengine' vendor ecosystem. - [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface (Category 8) typical of coding assistants that process untrusted external data.
- Ingestion points: Codebase contents are analyzed using
Grepand file-reading logic (e.g., intemplates/service.md). - Boundary markers: Absent; there are no specific instructions to treat processed code as potentially untrusted data.
- Capability inventory: The skill templates include file system writes (
Bun.write), network operations (fetch), and agent spawning (TeamCreate). - Sanitization: Absent; the skill relies on the underlying model's safety guardrails when processing discovered code.
Audit Metadata