skills/editframe/skills/agent-panel/Gen Agent Trust Hub

agent-panel

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: Includes ripgrep (rg) commands for searching the local project repository to identify event dispatchers and element schemas.- [PROMPT_INJECTION]: The skill describes a system (EFAgentPanel) that captures GUI state and element HTML to construct prompts for a coding agent, which represents an indirect prompt injection surface. Evidence: Ingestion points: ef-edit events containing element labels and HTML; Boundary markers: Not mentioned in documentation; Capability inventory: Aggregation of GUI edits for local prompt generation; Sanitization: Refers to the use of cleaned outerHTML snippets.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 05:50 PM