aidesigner-frontend

Warn

Audited by Gen Agent Trust Hub on Apr 18, 2026

Risk Level: MEDIUMDATA_EXFILTRATIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [DATA_EXFILTRATION]: The skill is designed to read and transmit comprehensive repository context—including framework configuration, Tailwind CSS tokens, existing components, and project routes—to a remote third-party MCP server at https://api.aidesigner.ai/api/v1/mcp.
  • [EXTERNAL_DOWNLOADS]: The skill fetches and executes the @aidesigner/agent-skills package from the npm registry using npx for setup, diagnostics, and porting guidance.
  • [COMMAND_EXECUTION]: The skill requires the execution of shell commands such as npx @aidesigner/agent-skills init, doctor, and adopt to manage its lifecycle and process generated designs.
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface.
  • Ingestion points: Repository content including framework files, UI components, and the DESIGN.md documentation file.
  • Boundary markers: Absent; the skill instructions do not specify any delimiters or ignore-instructions for the ingested repository data.
  • Capability inventory: Remote tool execution (generate_design, refine_design), local file system writes (.aidesigner/runs/), and shell execution (npx).
  • Sanitization: None specified; the skill passes repository context directly to the remote engine for processing.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Apr 18, 2026, 12:07 PM
Security Audit — agent-trust-hub — aidesigner-frontend