What problem does it solve? Research sources arriving from Drive folders, discovery, feeds, and queued URLs pile up in a notebook's _inbox/pending/ queue with no clear owner. This Skill sorts each source into the right intake bucket so the correct reader ingests it, eliminating duplicate fetches and unsorted backlogs. ## Core Features & Use Cases - Bucket-based routing: Reads the bucket definitions (folder, purpose, reader) from _inbox/intake/README.md and moves each pending ticket into exactly one bucket via wiki-triage.py route, with every call logged as precedent. - Raw capture for other readers: Fetches and stores the raw content of any source destined for another reader's folder, so no one re-fetches; login-gated pages are captured through the user's browser session. - Reader notification: Reports one table per folder to the user and posts a single tagged line per reader in the project's Discord channel when one exists. - Use Case: A shared research notebook (agentic-design) accumulates 20 queued URLs overnight. Run triage to file each into main, agent-builder, or ai-money buckets with one-line reasons, then notify each project reader of their new items. ## Quick Start Ask the assistant to triage the pending intake queue for your notebook and route each source to the right bucket.
