chromatic-troubleshoot-diff

Pass

Audited by Gen Agent Trust Hub on Sep 29, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill functions as a specialized troubleshooter for Chromatic visual regressions. It uses a clear, evidence-based diagnostic workflow and provides standard ripgrep search commands to help the agent locate relevant configuration and code patterns within a project.
  • [INDIRECT_PROMPT_INJECTION]: The skill exhibits a common attack surface for indirect prompt injection by ingesting untrusted data such as code snippets and build information from the user.
  • Ingestion points: Users provide build URLs, story names, screenshots, and story code snippets for analysis.
  • Boundary markers: The skill mandates a structured 'Diagnosis Card' output format, which helps maintain operational boundaries, though it lacks explicit instructions to ignore embedded instructions in the provided snippets.
  • Capability inventory: The agent is instructed to perform search operations (ripgrep) on the repository and read local configuration files (package.json, .storybook/preview.*). It does not have file-write or network-exfiltration capabilities.
  • Sanitization: No explicit sanitization or filtering of user-provided content is mentioned, but the diagnostic focus and restricted output format minimize the potential for harmful exploitation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 29, 2026, 09:46 AM