Pipedrive drifts where the reps move fast. Deal titles get named three different ways, notes pile up half-finished after a quarter, organization records go stale, and the custom fields you report on never quite line up. Cleaning it up by hand means opening one deal at a time, right when you need the pipeline clean for a quarter-end report or a sales-team handoff.
Scratch pulls your Pipedrive deals, persons, and organizations down as files on your computer. Your AI reads and edits every record, not the first few hundred, about 10x faster than it works over an API because it reads the files directly instead of making a call per record. Every change comes back as a word-level diff next to the original, and nothing writes back until you approve it, per record. Scratch is the companion app to Whalesync: clean the pipeline up here, then let Whalesync keep it in sync once the records are right.
What Scratch edits in Pipedrive
- Deal titles and custom fields
- Person and organization names, notes, and custom fields
- Deal stage assignment (the AI can move deals between stages if you ask)
How it works
- Scratch pulls your Pipedrive into files. Deals, persons, and organizations come down to a folder on your laptop, one record per file. Owners, timestamps, monetary values, and currency are pulled for context only and stay read-only.
- Your AI edits the fields you point it at. Open the folder in the agent you already use. The AI rewrites the prose fields, the deal titles, notes, and names, and updates custom fields. It can also move deals between stages if you ask it to. Try a prompt on a few records, then let it run across the whole pipeline. The AI edits the files, never the live CRM.
- You review every diff and write back. In the Scratch desktop app, each changed field shows next to the original, word by word. Validators enforce field length and any never-touch rules so a tightening pass does not drop a field your reps depend on. Approve what looks right, and Scratch writes only the records and only the fields you approved back through the Pipedrive API. Untouched fields are left exactly as they were.
What teams use it for
- Standardize deal titles across the pipeline so the board reads consistently.
- Tighten and normalize deal notes after a quarter of drift, before a quarter-end report.
- Clean up person and organization records ahead of a sales-team handoff.
- Normalize the custom fields your forecast and reporting depend on.
- Reorganize deal stages in a reviewed pass when the pipeline structure changed.
- Reconcile records after a migration into Pipedrive, so the source of truth matches again.
Why not let AI write straight to Pipedrive?
A direct API write or an MCP server hands the AI the publish button straight to your live CRM, and it is the slow path on top: every record is an API call, and on a real pipeline it falls over. There is no diff, no review queue, no rollback. One confident bulk pass rewrites every record at once, fires the automations watching those fields, and skews the forecast and the reports your team reads on Monday. By the time you catch the bad edit it is already live, and the original is often gone.
Scratch gives the AI the same access, but against a local copy. The write-back step is pulled out and handed to you. The AI can change anything in the copy; only you commit it, you keep the publish button, and Scratch writes back only the fields you changed. Every published record is reversible on its own, so a quarter-end cleanup is never a one-way door.
What's safe, and what's locked
Owners, timestamps, monetary values, and currency are pulled for context only and stay read-only, so a tightening pass cannot touch the numbers your forecast depends on. You bring your own AI: Scratch holds no AI credentials and runs no model, so you sign into Claude, Claude Code, Codex, Cursor, Copilot, Cline, or Windsurf the way you already do. Nothing leaves your machine until you publish. Scratch writes back only the fields you changed, so untouched fields are left exactly as they were, and locked or never-touch fields you flag are stripped before write-back. Validators enforce field length and never-touch rules so a tightening pass does not drop a field your reps depend on. Every published record is reversible per record, and the optional Python validators can be authored by the AI. Scratch is SOC 2 compliant.
Questions Pipedrive teams ask
Will a bulk pass fire automations or skew my forecast?
Edits in the local copy fire nothing and change no number. Owners, monetary values, and currency stay read-only, so the forecast inputs are not touched at all. When you approve a record, Scratch writes it back like any normal update, so an automation watching that field can fire, the same as a manual edit would. The difference is that you choose which records write back, instead of one bulk pass tripping every automation at once.
Can the AI move deals between stages?
Yes, if you ask it to. Deal stage assignment is editable, so a reviewed pass can reorganize the pipeline when the structure changed. You still see every move as a diff and approve it before it writes back. Owners, timestamps, and values stay read-only throughout.
Will a tightening pass drop a custom field my reps depend on?
No. Validators enforce field length and any never-touch rules, so a field your reps depend on is flagged next to the diff before anything writes back. Scratch also writes back only the fields you changed, so untouched custom fields are left exactly as they were.
Can I undo a run after it wrote back to Pipedrive?
Yes. Every written record is reversible from Scratch, per record. Roll back one deal or the whole run, and Scratch restores the original for you to publish. A quarter-end cleanup is never a one-way door.
See it on your own Pipedrive
Pull a few hundred records to feel the loop, then point the agent at the whole pipeline.