WAH doesn't start with a task list. It starts with the work graph. A line called Fix authentication tells you what to do — the graph tells WAH whether you should do it now.
That graph is why Ask WAH can answer What should I work on today? with the same confidence as What's blocking the launch? The competitors can draw the graph now — GraphClaw, Atlassian Teamwork Graph, Cortyx all do. WAH's moat is turning that graph into a moment-specific decision.
What the graph actually is
In lib/wah.ts, WORK_GRAPH is one task with a goal, a project, a due slice, and three context nodes: PR #284 on GitHub, a 2 PM call on the calendar, 4 messages in Slack. Recommendation: Fix authentication — Blocks onboarding and your 2 PM call is approaching. 35 min · High impact.
That tiny graph is the shape of every decision we rank. But not every task needs every node. Some are just Task → Project → Goal. Others grow: Task → GitHub PR → Slack conversation → Calendar event → Dependency → Launch deadline. The graph grows as WAH learns the context around the work — organic, not a schema you fill out.
WAH doesn't sort tasks — it routes from the graph
WAH does rank — rankNext() exists. The distinction is what gets ranked and why. A normal prioritizer: Task A → 9.2, Task B → 8.7. WAH: Task B, 35 min, Fits your window, Why this? → Unblocks Product Brief → on Launch path → you own the unblock → Friday deadline. Same math, different evidence.
When you dump what is on your plate in /app/onboarding, parseCapture() extracts work without required fields. rankNext() then filters to canAct and scores: leverage(task) + criticalPath(task) + urgency(task) + timeFit(task, window) + ownership(task) + momentum(task) − blockerPenalty − contextSwitchCost − partialWorkPenalty. The winner isn't the highest score alone — it's the decision with sufficient evidence. Orange must be earned.
The clock is the moat
The graph alone isn't enough. A competitor can build Goal → Project → Task → Dependency. WAH combines WORK GRAPH + CURRENT CONTEXT + AVAILABLE TIME + HUMAN OWNERSHIP = NEXT MOVE. Same graph, different decision:
90 minutes → Interview notes, 35 min → Do this next. 20 minutes → Schedule interviews, 20 min → Do this next. Waiting for legal → Pricing one-pager → You can wait. 2 minutes before a meeting → Send customer follow-up → Do this next.
Asana already considers tasks, deadlines, priorities and availability; ClickUp markets AI prioritization on deadlines, dependencies, goals. Table stakes now. WAH turns all of that context into one decision you can act on right now.
Why the graph is valuable
The graph is not valuable because it can be drawn. It is valuable because WAH can use it to make a better decision. That is harder to copy with marketing than a node diagram.
Screenshot the chain Interview notes → Product brief → Launch decision → Friday. The edge is real, but the value is that WAH followed it to choose Interview notes at 90 min and Schedule interviews at 20 min — same graph, different window, different answer.
WAH Todo is not a list with a model on top. It is a graph you can read, fix, and trust — then act on. The card is orange because the evidence is real — WORK GRAPH → CONTEXT NOW → RANK / ROUTE → DECISION EVIDENCE → DO THIS NEXT → START.
Your tools tell you what exists. WAH tells you what deserves your next 45 minutes — and why.
See the graph demo →