Skip to main content
The workflow catalog describes individual department procedures. The playbook layer describes how those procedures cooperate across department boundaries. A playbook is a governed journey that chains existing workflows (referenced by their stable XXX-NN codes, such as MKT-01 or SAL-01) into an end-to-end operating contract. The playbook layer does not replace the workflow registry, department packages, or client-facing workflow kits. Existing workflow IDs stay unchanged. Playbooks sit on top of them and make the transitions between owners explicit.

What a playbook defines

Every playbook carries the same governed contract:
  • Trigger and completion criteria: the event that starts the journey and the observable state that ends it.
  • Required inputs and expected outputs: the data the journey needs and the artifacts it must produce.
  • Primary owner and approval owner: who runs the journey and who signs off.
  • Handoff contracts: the ordered transitions between workflows (see below).
  • Stop conditions: states where the journey must halt, such as a lead opting out or a missing payment term.
  • Evidence requirements: the records that must exist to prove the journey ran correctly.

The five canonical playbooks

The initial layer seeds five high-value journeys covering the revenue spine, finance controls, client continuity, and the feedback loop into marketing. Department-only playbooks can be added later without changing the handoff model.

Handoff contracts

A handoff contract is the unit that governs each transition inside a playbook. Each handoff records:
  • From and to workflow codes: for example, MKT-01 hands off to SAL-01.
  • From and to department codes: makes cross-department boundaries explicit.
  • Trigger signal: the condition that fires the handoff, such as “Lead is confirmed as MQL and ready for a commercial conversation.”
  • Required payload: the fields the receiving owner must get, such as lead identity, fit evidence, offer interest, and next action.
  • Receiving owner: the person accountable for accepting the handoff.
  • Approval requirement: whether the handoff needs sign-off before it proceeds.
  • Fallback protocol: the manual path when an integration or automation is unavailable.
  • Stop condition: when the handoff must not proceed, such as a rejected fit or missing required fields.
  • Evidence requirements: the records that prove the handoff happened.
The fallback protocol is mandatory. A playbook must remain operable when an integration, paid tool, or automation endpoint is unavailable. The receiving owner must still be able to accept the handoff manually and preserve evidence.

How playbooks relate to workflow kits

  • The workflow kit catalog packages individual workflows as self-contained kits, one per XXX-NN code.
  • Playbooks reference those same codes in their handoffs, so MKT-06 in a playbook means the same Content Creation workflow as the MKT-06 kit.
  • Kits answer “how do I run this workflow?” Playbooks answer “what happens next, who receives it, and what must travel with it?”
Reference workflows by their stable IDs in handoffs, commit messages, and change control entries, exactly as the catalog does.

Viewing and seeding playbooks

The app exposes playbooks in two ways:
  • The Playbooks page in the app UI lists the seeded journeys with their handoffs.
  • The API serves them at GET /api/playbooks and GET /api/playbooks/:id. See Playbooks endpoints.
Seed or refresh the canonical layer from the repository root:
The seed script is idempotent. Re-running it upserts the five canonical playbooks by ID, marks them active, and rewrites their handoffs in order, so you can run it safely after pulling new playbook definitions.