Overview
The Pipefy Integrations engine has received an update that strengthens three areas: credential security, stability during high-volume operations, and an optimized build and debugging experience.
What's new?
- Variables Manager: your credentials are now stored as encrypted variables and referenced inside your steps. Values like `client_id`, `client_secret`, and tokens no longer appear in logs or in the Code step.
- Better handling of large flow runs: when a flow produces a very high volume of data, that output is now stored outside the main database. Heavy flows keep running smoothly, with no slowdown.
- Brand-new testing panel: a redesigned testing experience with Split and Drawer modes, Friendly View outputs, and one-click step testing right from the Canvas.
- A more stable engine, with clearer error messages: when a step fails, you get a clear description of the error and a shortcut that takes you straight to the step that broke.
How does this help?
- Credentials out of logs and Code: secret exposure is no longer a risk, a direct win for governance and compliance in Enterprise accounts.
- Reliability at scale: fewer crashes on large flows, high-volume operations keep running without interruption.
- Faster problem resolution: when something breaks, you see exactly what happened and jump straight to it, no hunting for the source of the error.
- Visual debugging: testing step by step on the Canvas turns building into something guided, instead of trial and error.
Use cases
1. A Pipe Admin connecting an ERP to an accounts payable process.
The connection uses the finance system's `client_id` and `client_secret`. With the Variables Manager, those credentials are encrypted as variables and referenced inside the steps without ever appearing in execution logs or in the Code step. The security team can approve the integration with no plaintext secret exposure to flag, and the account audit no longer has that open item.
2. A Builder syncing a high volume of records between Pipefy and a data warehouse.
The flow pulls thousands of rows on each run. Before, that volume strained the engine. Now the heavy output is stored outside the main database, the flow completes the sync with no slowdown or crash, and the nightly data routine runs to completion predictably.
3. An operations analyst debugging an integration flow with a ticketing system.
While building the logic on the Canvas, they test each step with one click and read the output in Friendly View, in Split mode, comparing input and result side by side. When a transformation step returns an empty field, they pinpoint the exact spot right away, without publishing the whole flow just to find where the problem was.
Most relevant for
- Pipe Admins
- Builders
- Operations teams
- IT Leaders
Plans and availability
This update is rolling out gradually. Part of the base already has access, and release to the remaining environments follows each one's upgrade window.
