Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/xxyoudeadpunkxx/decision-rain-library-project/llms.txt

Use this file to discover all available pages before exploring further.

Decision Rain Library Project uses five named Raindrop sub-collections nested under a single parent collection. The naming and hierarchy are intentional — the AI assistant reads collection names as lifecycle signals. When a link sits in 00_INBOX, the assistant knows it is unprocessed. When it sits in 20_LIBRARY, the assistant knows a human operator explicitly validated it. Renaming or reordering collections breaks those signals.

​
Collection Layout

Create the following structure in Raindrop. The parent is a regular collection; the five sub-collections live directly inside it. 
Decision Rain Library Project
  00_SYSTEM
  00_INBOX
  10_REVIEW
  20_LIBRARY
  90_ARCHIVE
Create these names exactly. Do not abbreviate, reorder, or add extras until you understand how the assistant uses them.

​
Collection Roles

Canonical rules, templates, tag registry, and golden examples. The assistant reads this collection before doing any serious classification work.This collection never contains links you found online. Every entry here is a governance document stored as a Raindrop bookmark using a stable placeholder URL. The bookmark’s excerpt or note field is the source of truth — not the URL.
Links the assistant has analyzed and written a proposal for, but the operator has not yet reviewed. Entries here are proposals, not decisions.A link may stay in 10_REVIEW indefinitely while you think it over. Nothing moves out without operator action.
Evaluated and operator-validated decision entries. This is the canonical collection. Nothing lands here without your explicit approval — not the assistant’s confidence, not clean formatting, not strong source evidence.An entry in 20_LIBRARY means: a human operator reviewed the proposal and said yes.
Rejected, outdated, duplicated, inactive, or historical entries. Kept for traceability. Archiving an entry requires operator approval just like promoting one — the assistant may propose archival but cannot act without your sign-off.

​
The State Machine

Collections are lifecycle states, not topical folders. Every link follows this transition path: 
signal → 00_INBOX → analysis proposal → operator validation → 10_REVIEW / 20_LIBRARY / 90_ARCHIVE
The assistant moves a link forward only after you explicitly approve the destination. The typical paths are:
  • Promote to LIBRARY — you reviewed the proposal and the entry is ready to be a canonical decision.
  • Send to REVIEW — the proposal is reasonable but you want to revisit it before committing.
  • Archive — the link is rejected, a duplicate, or no longer relevant.
A direct write into 20_LIBRARY is only valid after explicit operator validation. Previous similar approval, high assistant confidence, clean formatting, or strong source evidence do not count as validation. If the operator has not said yes for this specific entry, it stays in 00_INBOX or 10_REVIEW.

​
Why the Naming Matters

The numeric prefixes serve two purposes:
  1. Sidebar ordering — Raindrop sorts collections alphabetically. Prefixes like 00_, 10_, 20_, and 90_ force the collections to appear in lifecycle order in the sidebar, not random alphabetical order.
  2. Assistant signals — The assistant reads collection names to determine lifecycle state. 00_INBOX means unprocessed. 20_LIBRARY means validated. These are not conventions the assistant infers; they are names it is trained to interpret directly from the SYSTEM documents.
If you rename a collection — for example, changing 20_LIBRARY to LIBRARY — you must update every SYSTEM document that references the old name before using the system again.
Never skip INBOX. Even if a link seems obviously ready for LIBRARY, it must pass through 00_INBOX first. Skipping INBOX removes the trace and breaks the state machine. The assistant cannot safely propose a verdict on a link it was never seen to receive.

Build docs developers (and LLMs) love

Get started for freeTalk to us