You already know the drag: waiting through sequential steps, context-switching between terminal, IDE, and chat, watching work disappear into someone else's rented sandbox. ShipSpace runs the whole loop — research, implementation, review, verification — concurrently against a real directory on your disk. Real files back on disk, a decision-ready verdict at the end, keys stay in your keychain.
A controlled environment where structured build work becomes real output — with a verdict at the end of every mission.
Every run is a mission with an explicit goal, a working directory, context sources, and a roster of roles. Start from a preset — security audit, feature build, code review — or define your own, and reuse it without rebuilding the prompt each time.
Research, orchestration, implementation, review, and verification run concurrently across your codebase. One mission covers what would otherwise be five sequential sessions in five separate tools.
Per-role output streams, file change summaries, and inline editing via Cmd+P Quick Open. See what each role produced and what landed on disk without leaving the workspace.
Runs natively via Tauri v2 on your local filesystem. Preserves your shell config, keeps your API keys in the OS keychain, and writes real files to disk — nothing is rented, nothing leaves your machine.
Every mission ends with a decision-ready summary: what passed, what failed, what changed on disk, what to revise. Know in one screen whether to ship, iterate, or start over.
Keep the mission, the directory, the context, and the prior verdict — then launch the next run in one click. The iteration loop is a first-class action, not a cold restart.
Six starting points that map to the build work independent teams actually run. Pick one, point it at a directory, and ship.
Dispatch a mission across your codebase to surface unsafe patterns, secret leaks, and risky dependencies. Returns a prioritized issue list with file locations, severity, and a verdict on what to fix first.
Hand over a mission spec, a directory, and context. Get a working feature back — implementation, tests, and a verdict on what's ready to ship versus what needs revision.
Run a reviewer role across a branch or directory. Returns a structured review — what's solid, what's risky, what to change — mapped to specific files and lines.
Point a mission at a failing test or reported issue. The roster reproduces, diagnoses, fixes, and verifies — with changed files and a before/after summary at the end.
Launch a refactor mission with an explicit scope. Parallel roles handle analysis, implementation, and verification — so the change lands safely, not speculatively.
Send in a perf mission with a baseline and a target. Get profiling output, bottleneck analysis, and a verdict on whether the change actually moved the number.
Native shell processes
Drag & drop workspaces
WebGL GPU rendering
OSC 133 integration
ShipSpace runs natively on your OS, operates on your local filesystem, and keeps your API keys on-device. Missions produce real files on disk you can open, diff, and ship — not artifacts trapped inside a rented sandbox.
How the execution engine fits into the rest of MakeShipHappen.
ShipMind helps you understand the work. ShipSpace helps you execute it. The handoff is explicit — mission context moves cleanly from the second brain to the engine.
Ingest notes, research, and source material in ShipMind. Ask grounded questions. Generate build-ready artifacts.
Send mission context — goal, directory, constraints — into ShipSpace. The handoff is explicit, not implicit.
ShipSpace runs the mission across roles: research, orchestration, implementation, review, verification.
Inspect outputs, read the verdict, open the changed files, decide what ships. Iterate without losing context.