write-script-bun
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent on using the
wmillCLI tool for tasks such asscript preview,script run, andgenerate-metadata. These are standard development operations for the Windmill ecosystem. - [DYNAMIC_EXECUTION]: The
windmill-clientSDK described in the skill includes capabilities for programmatically executing scripts and flows within the workspace via methods likerunScriptByPathandrunFlow. - [INDIRECT_PROMPT_INJECTION]: The skill documents preprocessor scripts that handle untrusted data from external triggers (e.g., webhooks, emails). This functionality inherently creates a surface for indirect prompt injection if the data is not sanitized before use in downstream agent tasks.
- Ingestion points: Preprocessor scripts receive a single
eventparameter containing data from external triggers (SKILL.md). - Boundary markers: The instructions do not recommend specific boundary markers or delimiters for isolating external trigger content.
- Capability inventory: Scripts can use the
windmill-clientSDK to trigger other jobs, manage storage, and access sensitive resources (SKILL.md). - Sanitization: No explicit requirements for data validation or sanitization are provided in the instructions.
Audit Metadata