What problem does it solve? Manually rating the long-term importance of a large Todoist backlog is tedious and error-prone, especially when scores must be written back to a SQL database without mismatches. This Skill structures the process into batches of seven tasks, records David's 0-10 importance scores against explicit todoist_id values, and keeps companion data stores in sync. ## Core Features & Use Cases - Batch Scoring Workflow: Presents 7 unscored tasks at a time (lettered A-G) with progress tracking, then saves each score via explicit SQL UPDATE statements keyed by todoist_id. - Voice-to-Text Safeguards: Handles transcription artifacts like repeated decimals ("8.8" rounds down to 8) and duplicate confirmations. - Database Integration: Reads and writes docs/data/todoist-task-dimensions.db, optionally fetches fresh tasks from the Todoist API, and updates user-behavior.db and brain/ notes. - Use Case: David says "score my inbox tasks" and the assistant pulls unscored tasks from the SQLite database, presents them seven at a time, records his spoken scores, and reports the final distribution with the top 20% ranked. ## Quick Start Ask the assistant to score the importance of your unscored Todoist tasks in docs/data/todoist-task-dimensions.db, then reply with a 0-10 score for each lettered task.
