Choreseed: Make the chore leave behind a routine

Description and use

Choreseed creates reusable routines to clean contact CSVs without repeated AI calls, supporting offline replay and recipe storage in IndexedDB.

Problem

Cleaning a contact CSV without writing code and avoiding repeated AI calls for each file

Outcome

Zero new inference calls on the second compatible file; routine can be replayed offline with cached app and saved recipe

Workflow

• Collect instruction: "Trim Name and Email. Lowercase Email. Remove completely empty rows and duplicate contacts."
• Send instruction, column names, clarification, and selected reference excerpts to Fastify
• Backboard calls Qwen3.5-35B-A3B to generate a typed recipe
• Validate recipe with Zod and semantic checks
• Save recipe to IndexedDB
• Load saved routine and choose a compatible CSV
• Run routine locally in browser Web Worker using PapaParse
• Export new CSV
• Replay routine on new file without additional AI calls

Evidence

Choreseed turns ordinary instructions and selected guidance into a routine you can inspect, rehearse, save, and replay.

Source

The first release handles selected-field trimming, explicit case changes, column renaming, completely empty rows, and deduplication with an explicit key and keep rule.

Source

The recording starts with this instruction: Trim Name and Email. Lowercase Email. Remove completely empty rows and duplicate contacts.

Source

After review, six fictional input rows become four output rows. Leading-zero IDs stay strings. Two rows with missing emails survive separately. The original file stays untouched.

Source

The saved routine handles it with zero additional AI calls. A changed header set is blocked, rather than silently remapped.

Source

The free preview is usable for the fictional sample, editing, CSV export, recipe import/save and compatible-file replay, including offline replay after caching.

Source