Orca
Coordinate coding agents in separate worktrees, compare their changes, and feed review findings back into bounded tasks.
- Development Workflows
- Multi-Agent Collaboration
- Code Review
- GitHub
Stably: Orca and Parallel Agent Work
Publishes Orca, a workspace for parallel coding agents, isolated worktrees, and code review.
About Stably
Stably is the public GitHub organization publishing the canonical Orca workspace. Orca combines existing CLI agents, repository worktrees, terminals, and review interfaces. Its source makes the coordination layer inspectable without turning the external agent engines into Stably-owned models.
For a Floatboat workflow, the useful pattern is to define bounded assignments, isolate changes, and compare evidence before accepting an implementation. Parallel agents still need a shared task contract and an explicit integration owner. Running in separate worktrees does not itself prevent mistakes, conflicts, or unauthorized actions.
Stably Projects
Orca coordinates agent work with isolated file changes and review before integration. Explore its workflow model through task ownership, fixed revisions, and shared checks, keeping worktree isolation separate from the execution permissions and final merge decision.
Explore Other Publishers
Meet other people and teams publishing open tools, projects, and Agent Skills.
Stably FAQ
Which project is attributed to Stably here?
Orca is published in stablyai/orca, which is the upstream selected for this profile.
Why not use another repository named Orca?
The user described a widely used multi-agent workspace; its functionality and dated GitHub evidence match that description. Other same-name tools are separate works.
Does Stably publish the agent engines inside Orca?
No. Orca coordinates external engines such as Codex and Claude Code. Those engines retain their own publishers and accounts.
What makes worktree coordination useful?
It separates task files and supports comparing changes, provided scope, review, and final integration are explicitly owned.
Does MIT include model subscriptions?
No. MIT applies to the published project source. Provider accounts, model usage, and remote services are separate.
Where should release and support details be checked?
Use the canonical repository releases and the product documentation linked from its README. Avoid an unrelated package just because it has the same name.
Review One Orca Result Before Integration
Check engine authentication and remote-access settings in the Orca guide, then test an accepted change in the combined revision while keeping the merge decision explicit.