Your First Flow
Creating a Flow
Section titled “Creating a Flow”A Flow represents any unit of work — a feature, bug fix, or task. To create one:
- Open your project
- Click “New Flow” or use the + button on the Kanban board
- Enter a summary (short title describing the task)
- Optionally add a description and acceptance criteria
- Choose a type: Feature, Bug, or Chore
- Click Create
Your Flow starts in the idea state.
Moving Through States
Section titled “Moving Through States”From Idea to Planning
Section titled “From Idea to Planning”When you are ready to work on a Flow, move it to planning. If you have an AI agent connected, it will:
- Analyze the requirements
- Research the codebase
- Create a detailed implementation plan with tasks
Approval
Section titled “Approval”Once the plan is ready, the Flow moves to approval. You can:
- Read the plan — Review the proposed approach, file changes, and acceptance criteria
- Approve — Move the Flow to ready for implementation
- Reject with feedback — Send it back to planning with your comments
Implementation
Section titled “Implementation”When you (or the AI agent) start implementing, the Flow moves to in_progress. During this phase:
- Code is written following the approved plan
- Tasks are checked off as they are completed
- Git commits are tracked and linked to the Flow
Review
Section titled “Review”After implementation, the Flow moves to review. You receive:
- A summary of what was done
- Testing instructions — How to verify the changes
- Links to the PR and commits
You can then:
- Test the changes yourself
- Approve — Move the Flow to done
- Reject with feedback — Send it back to in_progress
The Flow is complete. The code is merged, time is tracked, and everything is documented.
Flow Details
Section titled “Flow Details”Each Flow has a detail view showing:
- Description and acceptance criteria
- Implementation plan (after planning)
- Tasks with completion status
- Git information (branch, commits, PR)
- Agent sessions (AI work history)
- Time entries linked to this Flow
- Write clear acceptance criteria — they help the AI agent understand what “done” means
- Review plans carefully — a good plan leads to better implementation
- Use the Kanban board for a visual overview of all your Flows