The Full Idea Pipeline
Capture → triage → classify → convert. The five-stage flow that keeps ideas moving and rejects the ones that shouldn't.
By the end of this module
The Five Stages
Every idea passes through this pipeline, even the ones you in the end archive:
Capture → RAW → IN_TRIAGE → CLASSIFIED → (CONVERTED or PARKED or ARCHIVED)
The discipline is flow rate. Ideas that sit in RAW or IN_TRIAGE for weeks are dead weight: they consume working memory, generate stale-capture alerts, and slowly convince you the system isn't working. The system is working, your routing is broken.
Target rates: nothing sits in RAW more than 14 days, nothing in IN_TRIAGE more than 7. If you can't hold those numbers, either capture less or process more, those are the only two levers.
Stage 1. Capture (30 seconds)
Cmd+Shift+K, title, type, save. The title alone is often enough. Resist the urge to write the full idea at capture; that's triage's job. Speed is the metric, anything that pushes capture past 60 seconds is friction you'll feel as a slow accumulation of uncaptured thoughts.
Stage 2. Move to Triage (5 seconds)
From /vault or /today, click "Start Triage" on a RAW idea. The idea transitions to IN_TRIAGE and opens the scoring panel.
The right cadence: batch triage during weekly review. Doing one idea at a time costs context-switching overhead; doing 5+ in a session gets you into a scoring rhythm where calibration improves across ideas. Weekly review Step 3 is built for exactly this.
Stage 3. Score 16 Criteria (3-5 minutes per idea)
All 16 criteria on a 1-10 scale. The system computes a weighted total automatically.
The single rule that matters: score honestly, not optimistically. A 3 on revenue potential for an unvalidated idea is honest. A 5 that's pure hope is just hope. The model rewards honesty mechanically, optimistic scoring just produces inflated totals that all look equivalent, which defeats the entire purpose.
Don't skip the notes field. Notes are how future-you (in three months, scoring a similar idea) calibrates against current-you's reasoning.
If you have Pro, AI Score pre-fills all 16 criteria as a starting point. Adjust freely, the AI is a second opinion, not the verdict. Use it especially when your queue has 10+ ideas and manual scoring fatigue would otherwise compress your scoring range.
Stage 4. Assign Outcome (10 seconds)
Pick the outcome that matches the score and your strategic context. The 10 outcomes encode different kinds of yes and no:
- Yes, now: BUILD_NOW, EXPLORE_ACTIVELY
- Yes, but not now: EXPLORE_LATER, COMPOUNDING_ASSET
- Yes, but wrong shape: EXCITING_BUT_DISTRACTING, STRATEGICALLY_STRONG_RESOURCE_HEAVY
- No, costly: MAINTENANCE_TRAP
- No, defer: PARK
- No, dead: ARCHIVE, KILL
The critical rule, repeated because everyone misses it: BUILD_NOW does not create a project. It's the system saying "if you have capacity, this is the strongest candidate." You still need to open Focus, check your capacity, and explicitly create the project. That step is intentional, triage tells you the opportunity is strong, only you can decide whether you have the room.
Stage 5. Convert to Asset (Optional, 2 minutes)
For CLASSIFIED ideas with capture types INSIGHT / PROMPT_ASSET / FRAMEWORK / DECISION / LAUNCH_ANGLE / MONETIZATION_IDEA, the Convert to Asset button appears.
Convert when the idea has reusable substance, a prompt you'll re-run, a framework you'll apply to multiple projects, a learning you'll reference. Pick a specific asset type (never default to OTHER). Mark it reusable if it applies broadly.
The CONVERT_TO_ASSET recommendation fires after 14 days if you have 2+ asset-eligible CLASSIFIED ideas. That's the system's nudge to either convert them or archive them, letting them sit indefinitely is the failure mode.
Pipeline Health Checks
Use these as monthly diagnostics:
- RAW count holding under 15? Capture/triage ratio is healthy.
- No ideas stuck in IN_TRIAGE >7 days? Triage habit is intact.
- CLASSIFIED count growing without being acted on? You're scoring but not deciding, review your Focus before scoring more.
- Assets section growing? Pipeline is producing payoff, not just throughput.
- PARK count growing without ARCHIVE? You're hoarding. Trigger a quarterly park-sweep.
Practice Checkpoints
Sign in to track your progress and mark checkpoints complete.