Vault. Your Compounding Library

Ideas are ephemeral; assets compound. The vault is where the difference shows up.

The Vault holds two distinct object types: Ideas (the inbox) and Assets (the library). Most users treat them as the same thing. They are not.

Ideas are observations and bets. They may or may not become real. Most won't. They have a lifecycle that ends in CONVERTED, PARKED, or ARCHIVED, they're not designed to persist as ideas forever.

Assets are intentionally preserved knowledge. A prompt you re-run every week. A pricing framework you've used on three launches. A "what we learned" document from a failed project. These are the things that compound.

Here's the math: an idea you used once gave you one unit of value. An asset you've reused 12 times has given you 12. A vault with 200 ideas and 4 assets is producing less uses than a vault with 30 ideas and 40 assets.

Ideas pass through these statuses:

RAW → IN_TRIAGE → CLASSIFIED → CONVERTED (terminal)
↓
PARKED → ARCHIVED (terminal)

You don't need to memorize this. The page shows it. What you need to know is what each state means in practice:

• RAW: Inbox. The system gently nags after 7 days.


• IN_TRIAGE: You opened the scoring panel. STALLED_TRIAGE fires if it sits here 7 days.


• CLASSIFIED: Triage complete. A score and outcome exist. Idea is now an evaluated opportunity, not a guess.


• CONVERTED: An asset was created from it. Highest-value end state for ideas with reusable substance.


• PARKED: Deferred without prejudice. Revisitable. REVISIT_PARKED fires at 60 days for high-score parked ideas.


• ARCHIVED: Closed. Not deleted, still searchable, but out of the active view.

An asset is worth creating if it answers yes to two questions:

1. Would I reuse this if I had a clean way to find it?
2. Did building this take more than 10 minutes of real thinking?

If both are yes, convert it. If the thinking took less than 10 minutes, it's an idea, not an asset.

The Convert to Asset button appears on CLASSIFIED ideas with these capture types: INSIGHT, PROMPT_ASSET, FRAMEWORK, DECISION, LAUNCH_ANGLE, MONETIZATION_IDEA. Pick a specific asset type during conversion, never default to OTHER. The asset types exist because the type determines how the asset surfaces in future searches and recommendations.

Once an asset exists, mark it reusable if it applies broadly. Reusable assets fire in the CONVERT_TO_ASSET recommendation for new ideas with similar capture types, the library starts to recommend itself.

Assets that aren't curated rot. Block 30 minutes once a quarter and do this:

Archive dead assets. Anything you haven't touched in 180 days and can't picture using next quarter. Be ruthless. The cost of a stale asset is search-result clutter, every time you look for something, dead assets clutter the result.

Merge near-duplicates. Two prompts that do almost the same thing? Keep the better one, link the other as REPLACES, archive it.

Re-tag the survivors. A consistent tag taxonomy is what makes the asset library searchable. If the same concept lives under three different tag names, pick one and migrate.

Link to projects. For every survivor, ask "what active project would benefit from knowing this exists?" Create a REUSABLE_FOR relationship. Now the asset surfaces on those project detail pages.

Defaulting to OTHER as asset type. OTHER is what you pick when the asset genuinely doesn't fit. Almost nothing genuinely doesn't fit. Picking the right type from the 13 specific ones makes the asset surface to the right contexts later.

Parking instead of archiving. Park is for "I might come back to this in 1-3 months." Archive is for "I'm done with this." Indecision shows up as a wall of parked ideas. Parking is not a way to avoid the archive decision, it's a way to defer a decision you're confident you'll make later.

Searching the vault page when you should use Cmd+K. Global search is faster and crosses object types. Use the vault page for filtered browsing (e.g., "show me all PARKED ideas tagged 'pricing'"), not for finding a specific known item.