A Vibestrate flow is a reusable recipe for supervised AI coding. It defines the phases, roles, model seats, approval gates, validators, and handoffs that turn one brief into a run you can watch and approve.
What a flow is for
A one-off prompt chain is gone when the chat ends. A flow is a file: a named process for the work that keeps coming back, like bugfix loops, deep refactors, quality arbitration, docs syncs, migrations, and security review.
It is also how Vibestrate avoids depending on one model. The flow defines the work and the crew can change, so the same process runs on frontier models, cheaper ones, local ones, or a mix.
What a flow can define
- Phases: plan, architect, implement, review, fix, verify, or custom steps.
- Roles: planner, architect, executor, fixer, reviewer, verifier, or custom roles.
- Approval gates: the moments where the human stays in control.
- Validators: commands and checks that decide whether work can continue.
- Handoffs: what each phase passes to the next model or reviewer.
- Metadata: tags, summaries, risk level, version, and author identity.
Pull, fork, publish
| Action | Command | What it means |
|---|---|---|
| Install | vibe flows hub install quality-arbitration | Install a verified flow and run it with your own local setup. |
| Install community | vibe flows hub install mira@deep-refactor | Install a community-authored flow under the author handle. |
| Publish | vibe flows hub publish my-flow | Share your workflow so others can inspect, fork, and run it. |
Sharing a flow does not move anything to the cloud
A shared flow is a recipe. Vibestrate still runs on your machine, uses your vendor CLIs, keeps your keys out of the orchestrator, and records the run locally. Pulling a flow gets you the process and nothing else.
The hub is where people share how they run a given kind of job. You bring the models, the keys, the codebase, and the approval.
