Execution History
View past Flow executions, conversations, and activity logs.
Execution History
Every time someone uses one of your Flows, Agents, or Apps, FormWise keeps a record of it. The History page, in the Monitor section of the sidebar, is where you go to review recent runs, browse conversations, and replay what happened.
Think of it as your activity log - a timeline of everything your Flows have done, newest first.
Rolling out
On workspaces with the new access model, this page is called Activity and gains two more tabs next to Runs: Admin activity (what your admins changed) and Tool changes (what changed in your Agents, Flows, and Apps).
[Screenshot: The main History page showing a list of recent runs grouped by Today, Yesterday, and Last 7 Days]
Runs vs. AI Assistant conversations
The History list mixes two kinds of entries:
- Runs of your Flows, Agents, and Apps. A form-style Flow produces one run every time someone submits it. These rows show the Flow's logo.
- AI Assistant conversations - your own chats with the in-app AI Assistant. These rows get a sparkle icon and are labelled AI Assistant.
Runs are tied to an end user - the person who triggered them. See Users for more on how that identity works.
What the list shows
The History page lists the 100 most recent runs across your workspace. Each entry includes:
- Snippet - The latest message the user sent, or the Flow's name when there is no message.
- Flow name - Which Flow, Agent, or App was used.
- Time - When the run started.
Filtering and search
When you have a lot of runs, filtering helps you find what you need quickly.
- Search - Type a Flow name, a snippet of what the user said, or the name or email of the team member who ran it. AI Assistant conversations are searched by title.
- Filter by Flow - The filter dropdown next to the search box scopes the view to a single Flow. Pick All results to clear it.
- Filter by AI Assistant - A dedicated filter option shows only your in-app AI Assistant conversations, separating them from end-user activity.
Results are grouped into time-based sections so you can scan them easily:
- Today
- Yesterday
- Last 7 Days
- Last 30 Days
- Older
Need to filter by end user or by a custom date range? Open a specific Suite and switch to its History tab. The per-Suite view adds an end-user filter and a Today / 7d / 30d / custom range picker, plus token and cost columns.
Inspecting a single run
Click any run to open it. You see a replay of the run the way the end user experienced it: the user's messages, the AI's responses, and any artifacts that were produced (long-form generated content, code blocks, etc.).
The replay is live - you can type a follow-up message and pick up where the run left off. Your follow-up runs the Flow again as you; it does not send anything to the original end user.
This is the right view when you want to understand what the user saw.
Step-by-step logs
To see what happened under the hood, open the run from analytics instead: on the Flows page, open a Flow's ... menu and choose View analytics (Agents have the same option), then click a run in the list. The run detail has two tabs:
- Conversation - The same replay of what the user saw.
- Steps - A waterfall of every node in the workflow. Each step shows whether it succeeded or failed and how long it took, so you can spot the slow steps at a glance. Expand a step to see the input it received, the output it produced, and any error it returned.
[Screenshot: A run's Steps tab, showing a waterfall of nodes with one step expanded to its input and output]
Debugging failed runs
Run details are one of your most powerful debugging resources. When a Flow misbehaves:
- Open the Flow's analytics (... menu, then View analytics) and find the run that went wrong.
- Switch to the Steps tab.
- Scan the waterfall for the first step with an error - that is usually where the problem started.
- Expand that step to see the exact input it received and the error it returned.
- Cross-reference with the Troubleshooting guide for common error patterns.
Once you have identified the broken node, jump back to the Flow's workflow builder, fix the issue, publish a new live version, and re-test. The next run will appear at the top of the History list.
Understanding latency and token usage
Each step in the Steps waterfall includes the time that node took. A Suite's History tab adds a Tokens and a Cost column for each run - hover over the token count for the breakdown:
- Input and output tokens - Tokens consumed by the prompts sent to the model, and tokens produced in its responses.
- Model - Which models handled the run (helpful when you mix providers in one workflow).
- Cost - Shown either in credits, in USD when the run used your own API key, or as a mix of both.
Looking at these numbers over time tells you which Flows are expensive to run, which prompts are getting too long, and where you might switch to a cheaper model.
AI Assistant conversations
Your chats with the in-app AI Assistant appear in History alongside your runs. Use the AI Assistant filter to list only those, and the search bar to find one by its title.
Opening an AI Assistant conversation shows the full back-and-forth, and you can keep chatting right where you left off.
Why history matters
The History page is more than just a log - it is one of the most valuable resources you have for improving your Flows over time. Here is how you can use it:
- Debug issues - When a Flow fails or produces unexpected output, check the run's steps to see exactly which step caused the problem and what data it was working with.
- Understand your users - See what inputs people are providing, what questions they are asking your chatbots, and how they are interacting with your Flows in the real world.
- Identify patterns - Spot trends like recurring failures, popular Flows, or common user inputs that you could optimize for.
- Improve your workflows - Use real usage data to refine your prompts, adjust your logic, and make your Flows smarter and more reliable.
The more you review your history, the better your Flows become.
A note on retention
Run records persist for as long as your Flow and organization exist. Deleting a Flow removes its runs. There is no automatic time-based purge of runs.