PeopleOps Console
Use this skill to solve PeopleOps Console tasks that require reconciling multiple record types into one JSON answer.
Workflow
- Read the task prompt and the answer template first.
- Open the PeopleOps Console at the supplied runner URL and log in with the provided credentials.
- Inspect only the workspace areas the prompt names:
- employee, leave, payroll, recruitment, document, message, policy, and audit views as needed
- Treat the prompt as a source-order hint, then resolve conflicts with this rule:
- prefer submitted or approved assignment-history records over drafts, superseded records, stale profile summaries, and case-summary-only text
- use direct packet, ledger, policy, or audit evidence over inferred status
- exclude adjacent audit events when the prompt asks for a narrower scope
- Fill every field in the template with the exact type and label it requests.
- copy enum values literally from the template
- keep arrays to the requested identifier values only
- sum totals directly from the relevant ledger when asked
- Return only the JSON object. No markdown, prose, or code fences.
Common decision patterns
- Onboarding closeout: approve only when leave and payroll evidence are both clean; otherwise keep the closeout on hold or open remediation.
- Folder and notice reviews: a missing required file or a defective formal notice blocks approval and usually requires remediation or reissue.
- Leave precedence: an approved current-period assignment overrides a stale profile summary when ledger, policy, and audit evidence agree.
- Recruitment reconciliation: derive selection from committee decision plus offer confirmation; put waitlisted and rejected candidate IDs in their arrays; use the recruitment ledger for the total cost.
- Payroll readiness: use the submitted payroll assignment, exclude draft records, and tie accrual readiness to the submitted record plus the batch evidence.
Output discipline
- Use the answer template as the schema.
- Preserve key order when practical.
- Do not invent synonyms for normalized business labels.
- If evidence is incomplete, keep searching the app instead of guessing.