approve-workday-tasks

Pass

Audited by Gen Agent Trust Hub on Sep 1, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the chrome-cdp CLI tool to interact with a running browser instance. This includes commands like find, click, snap, and nav. While these commands control the browser, they are scoped to the user's session and the specific purpose of managing Workday tasks. The skill also references a local recipe file workday-approve-and-return located in the user's config directory, which is used to batch browser interactions.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes task information from Workday (titles, worker names, amounts). However, the attack surface is minimal as the data is used for identification and display to the user. The skill implements a verification step, requiring a match between the preview data and the detail pane before approval, which serves as a manual validation layer.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 1, 2026, 08:52 AM
Security Audit — agent-trust-hub — approve-workday-tasks