What problem does it solve? Getting timestamped data from arbitrary CSV exports (wearables, spreadsheets, Pipedream/IFTTT workflows) into a Fulcra account normally requires hand-writing API calls, managing annotation definitions, and building your own dedup logic. This Skill maps CSV columns onto Fulcra annotation events with deterministic source IDs so imports are idempotent and re-runnable. ## Core Features & Use Cases - Flexible column mapping: Map timestamp, title, value, note, tag, and duration columns to Fulcra DurationAnnotation or InstantAnnotation events, with type coercion and unit support. - Three target modes: Import into user-defined annotation definitions, built-in Fulcra types like BodyMass, or generic bootstrapped definitions. - Idempotent imports with export: Deterministic source IDs plus client-side dedup readback make re-imports safe, and the export command round-trips annotations back to CSV for auditing. - Use Case: A user has a spreadsheet of daily body weights and mood scores. Bootstrap a Mood definition, import the weight CSV against the built-in BodyMass type, and verify what landed with a filtered export. ## Quick Start Ask the agent to import your CSV file into Fulcra by running a dry-run first, for example: import my weights.csv into Fulcra as BodyMass instant annotations with the date and kg columns.
