Playbooks

Permissions: playbooks.read, playbooks.write

Purpose

App route: /automations/playbooks A playbook chains existing workbooks. Steps can use severity-based run_if gates and map structured output from an earlier step into parameters for a later step. References follow the step/output pattern shown by the editor.

Prerequisites and data sources

Tabs and actions

Freshness and scope behavior

Workflow overview

Build and run

Create manually, import a bundle, or generate a draft through the AI interview.
Add steps in dependency order and select a workbook for each.
Map playbook inputs and previous structured outputs into workbook parameters.
Add explicit conditions for steps that should skip on healthy/informational results.
Validate missing references and cycles, then run manually.
Inspect per-step status, inputs, raw/structured result, skipped reason, and final result.
Schedule only after repeated manual success.

Export bundles inline referenced workbooks for portability. Review imported workbook commands because portability does not guarantee that target environments, permissions, or identifiers match.

Interpretation of results

Exports, history, scheduling, and integrations

No dedicated export, history, scheduling, or integration controls are documented for this feature page.

Safety and limitations

Safety

Every step retains the workbook’s runtime and write risk. A condition is orchestration logic, not an authorization boundary. Keep remediation and verification as separate steps, fail closed when required output is absent, and do not map secret-bearing output into logs or notifications.

Troubleshooting

Symptom	Resolution
Step skipped unexpectedly	Inspect `run_if` and the prior step’s normalized severity/status.
Mapping is empty	Confirm the producer emits the named structured key.
Playbook stops	Inspect the first failed step; downstream required dependencies do not run.
Imported playbook is invalid	Import referenced workbooks together and resolve duplicate/missing IDs.