ui-convert-ambiguity-resolver

Pass

Audited by Gen Agent Trust Hub on Apr 10, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is designed for static analysis and classification of project source files. It reads local files and updates a specific mapping file (index.json).
  • [DATA_EXPOSURE]: The skill accesses source code and project configuration files (project.json, index.json). This access is necessary for its stated purpose of identifying UI components and logic modules.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted source code as data for classification. The potential impact is limited as the skill's capabilities are restricted to updating a classification field in a JSON file and logging results. It lacks dangerous capabilities such as network access or shell command execution.
  • Ingestion points: Reads source code files and dependency metadata from index.json.
  • Boundary markers: Absent. The skill instructions guide the agent to analyze the code content directly using a set of classification questions.
  • Capability inventory: Reads local project files, updates the artifact category in index.json, and writes logs to logs/skipped.jsonl. No subprocess calls, network operations, or dynamic execution patterns were identified in the skill logic.
  • Sanitization: Not explicitly defined in the instructions, but the output is restricted to a predefined list of valid categories (component, page, layout, style, irrelevant).
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 10, 2026, 02:27 PM
Security Audit — agent-trust-hub — ui-convert-ambiguity-resolver