muapi-workflow

Pass

Audited by Gen Agent Trust Hub on Sep 1, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to perform package management operations, specifically using pip install --upgrade on a local workspace path to maintain the muapi-cli toolset.
  • [INDIRECT_PROMPT_INJECTION]: The skill exposes a surface for processing untrusted data by taking natural language descriptions to generate and discover workflows. While typical for AI-driven workflow tools, it presents an attack surface for embedded instructions in user prompts. \n
  • Ingestion points: User-provided strings for the --prompt argument in scripts/generate-workflow.sh and the --query argument in scripts/discover-workflow.sh. \n
  • Boundary markers: None identified; user prompts are interpolated into command arguments. \n
  • Capability inventory: The skill can create, edit, and execute multi-step AI generation workflows using the muapi command-line tool. \n
  • Sanitization: None detected in the provided scripts; the skill relies on the underlying CLI to process inputs.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 1, 2026, 11:15 PM