improve-ui

Pass

Audited by Gen Agent Trust Hub on Jul 26, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill enforces strong security boundaries by restricting the agent to read-only access of repository source code and explicitly forbidding the installation of packages, mutation of the working tree, or execution of formatters and scripts.
  • [PROMPT_INJECTION]: The skill reads and processes external repository content, such as source code and design documentation, which establishes a surface for indirect prompt injection. The risk is minimized by the skill's restricted execution environment and limited write scope.
  • Ingestion points: Repository source code and documentation files (e.g., DESIGN.md) as specified in the audit instructions.
  • Boundary markers: Absent; the instructions do not provide explicit delimiters to separate ingested data from agent instructions.
  • Capability inventory: File system read access for auditing; file system write access limited exclusively to the design-plans/ directory.
  • Sanitization: Absent; the skill does not define methods for validating or filtering the content of ingested repository files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 26, 2026, 12:44 PM
Security Audit — agent-trust-hub — improve-ui