data-tidylisted

Clean up messy data into a structured, validated table — from any source (a junk-filled .xlsx/CSV, a pasted email/markdown table, a Word table, or an Outlook .msg) to a clean .xlsx plus an audit/change report. Use when the user wants to "clean up this data", "tidy this spreadsheet", "normalise this list", "structure this messy export", "dedupe this", "standardise these dates / currencies", or "turn this into a clean table". (For getting data OUT of PDFs — tables, forms, scanned documents — use data-extract, not this skill.) Works INTENT-FIRST: it asks what the data is for and the expected output up front, then profiles → proposes a transform recipe → you confirm → applies it deterministically → reports every change and flags cells for review. Never mangles silently; transforms are computed locally by a deterministic engine. Standalone — no other toolkit required. NOT deal-document intelligence (lease abstraction, model review, comps) — that's out of scope.
moonlight-lupin/data-toolkit · ★ 0 · Data & Documents · score 72

Install: claude install-skill moonlight-lupin/data-toolkit

# Data Tidy Turn messy data — in whatever shape it arrives — into a clean, validated structured table (`.xlsx`) plus an **audit/change report**. The design is **intent-first** (ask the purpose and the wanted output before guessing) and **human-in-the-loop** (propose → confirm → apply → report). Transforms are deterministic and logged; nothing is changed silently. > **Self-sufficient & local engine.** Runs on its own — no other toolkit needed. All processing > (including OCR) happens on your machine: no cloud OCR, no external APIs, no third-party uploads. > Note the AI agent driving the skill does send whatever it reads into its context to your AI > provider — "never leaves the machine" is not claimed. See `../../PRINCIPLES.md#data-handling--pii-policy`. ## Workflow ### 0 — Intent (ask up front — this is what makes it economical) Before profiling anything, ask (a short `AskUserQuestion`): 1. **What's this data for?** (the purpose / where it's going) 2. **What output do you want?** — the target columns / shape, e.g. `Investor | Commitment (£) | Close date (DD MMM YYYY)`. A one-line sketch is enough. 3. **Any rules?** — a master list to validate against, required fields, format preferences. This pins the target, so the profile inspects *against it* and the recipe is proposed toward a known destination — not guessed. Fewer tokens, fewer wrong turns. ### 1 — Ingest The engine is **shared** at the plugin-root `scripts/` (with `data-extract`). Add it to the path, run from this