Reading a Workflow's Reliability Score Before You Trust It With Production Data
Every workflow now has a reliability score you can check before you put real data behind it, instead of scrolling through a raw run history.
Before you wire a workflow into something that matters, you can now check a single reliability score instead of scrolling back through a log of individual runs.
What it is
Open any workflow's reliability panel and you'll see a computed score for that workflow, pulled from its own dedicated endpoint. It's meant to answer one question quickly: has this automation actually been holding up, or has it been quietly failing and retrying more than you'd want to know?
Why a score instead of a log
A raw execution history tells you what happened on any single run, but it doesn't tell you, at a glance, whether a workflow is generally trustworthy. A score turns a pile of individual runs into one number you can check before flipping something from rehearsal into production use — no need to eyeball a list of timestamps and guess at the pattern.
Where to find it
The reliability panel sits directly in the workflow editor, next to the other diagnostic views. It's meant to be a quick check, not a deep investigation — for the "what actually went wrong" question, pair it with the dependency graph and the run-diagnosis tooling we're also rolling out.
If you've been hesitant to trust an older automation with anything customer-facing because you genuinely don't remember how often it's failed, this is the fast way to find out before you decide.