workflows-history
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the official
zapier-sdkCLI tool to perform workflow management tasks such as listing workflows, runs, and drafts. These commands are part of the vendor's standard SDK and are used for their intended purpose within the skill. - [INDIRECT_PROMPT_INJECTION]: The skill processes workflow input and output data retrieved from the CLI. This data could potentially contain untrusted content from external sources processed by a workflow.
- Ingestion points: CLI output from
list-workflow-runs,get-workflow-run, andget-trigger-runinSKILL.md. - Boundary markers: The skill uses markdown code blocks and requests the agent to parse JSON fields, but does not specify explicit boundary markers for untrusted content.
- Capability inventory: File system access and network operations are mediated through the
zapier-sdkCLI tool. - Sanitization: None specified, though instructions to summarize specific fields reduce the likelihood of arbitrary instruction execution.
Audit Metadata