write-flow

Pass

Audited by Gen Agent Trust Hub on Oct 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute wmill CLI commands such as wmill flow new, wmill flow preview, and wmill generate-metadata to manage the lifecycle of local flow definitions.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied summaries and descriptions to generate flow definitions.\n
  • Ingestion points: User input for path, summary, and description (SKILL.md).\n
  • Boundary markers: Not explicitly defined for user input interpolation.\n
  • Capability inventory: Shell execution via wmill CLI and file system writes for flow.yaml.\n
  • Sanitization: Not explicitly specified; relies on the underlying CLI tool's behavior.\n- [DYNAMIC_EXECUTION]: The skill guides the agent in generating flow definitions (flow.yaml) that may include executable JavaScript expressions in input_transforms.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 4, 2026, 12:47 PM