Planner

The Planner decomposes complex goals into concrete, actionable work. Give it an objective and it returns a structured plan: epics, tasks, acceptance criteria, suggested sequencing. It understands your team's strategy context, so plans reflect your actual priorities — not a generic template.

What the Planner does

When you assign a task to the Planner, it reads your existing strategy tree, knowledge atoms, and product context to understand where you are and what's already been decided. Then it generates a concrete execution plan — not just a list of bullet points, but a properly structured subtree with:

• Named epics and tasks with clear statements
• Acceptance criteria that define what "done" looks like
• Suggested sequencing based on dependencies
• Flags for tasks that need human review or external input

When it activates

The Planner is most valuable at the start of a new initiative:

• You have a new objective or key result and need to figure out what work it requires
• You have a solution defined but need it decomposed into epics and tasks
• You want a second pass on an existing plan to catch gaps or improve sequencing
• You're doing quarterly planning and want to turn OKRs into a ready-to-execute backlog

What it needs

• A parent node in the strategy tree (objective, key result, or solution) to anchor the plan
• The more context in your knowledge graph, the better — decisions, principles, and prior learnings all inform the plan
• Clear acceptance criteria or a description of the goal if the parent node is sparse

What it produces

• A structured subtree of epics and tasks under the assigned parent node
• Each task has: a statement, description, and acceptance criteria
• A summary comment on the task explaining the plan's rationale and any flags

Example: planning a growth experiment

You have a key result: "Increase trial-to-paid conversion from 18% to 25%." You create a solution node — "Redesign onboarding flow" — and assign it to the Planner.

The Planner reads your:

• Prior decisions about the onboarding flow (from atoms in the Wisdom tree)
• Product tree — what components and features already exist
• Code tree — what's in the codebase related to onboarding

It produces 4 epics: User Research, Design, Implementation, and Measurement — each with 3–5 tasks, acceptance criteria, and a suggested order. The Researcher and Coder can be assigned tasks directly from the plan.

Best for

• Breaking down solutions into epics and tasks
• Quarterly planning from OKRs to a ready-to-execute backlog
• Creating roadmaps from scratch when you have a goal but no plan
• Gap analysis on existing plans — "what are we missing?"

What it reads

• Your strategy tree — ancestors and siblings of the assigned node
• Wisdom atoms — especially DECISION and PRINCIPLE
• The product tree — features and components
• Related prior tasks and outcomes