gathering-ui

Pass

Audited by Gen Agent Trust Hub on Mar 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill orchestrates development workflows by executing local CLI tools including gw for repository management and uv run for visual regression testing, component auditing, and database seeding via the glimpse utility.
  • [EXTERNAL_DOWNLOADS]: The instructions reference @autumnsgrove/lattice, which is a verified vendor-specific resource belonging to the skill author's ecosystem. This is used for standard UI component integration.
  • [PROMPT_INJECTION]: The sub-agent dispatch mechanism in references/conductor-dispatch.md interpolates user-provided UI specifications ({ui_spec}) into prompts. While this presents an indirect prompt injection surface, the risk is mitigated by the multi-stage validation process and isolation of sub-agent contexts.
  • Ingestion points: Human-provided UI specifications enter the context in Phase 1 of SKILL.md and are passed to the Chameleon sub-agent.
  • Boundary markers: Prompt templates in references/conductor-dispatch.md do not utilize explicit XML tags or delimiters for user input interpolation.
  • Capability inventory: The skill possesses capabilities for file system modification (gw) and local command execution (uv run) across multiple phases.
  • Sanitization: No explicit sanitization or filtering of the UI specification input is performed before dispatch.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 28, 2026, 02:20 AM