Boomship/Intelligence
Your model is the engine. Boomship is the machine.
Bring Claude, GPT, Gemini or OpenRouter. Boomship gives them the context, the computer, the proof and the memory.
ClaudeGPTGeminiOpenRouter
Autothe Boomship layerBrain · computer · artifacts · receipts
Act 01Switch
Switch the model. Nothing else moves.
Mission 0218 · IllustrationLaunch our new product.
DecisionLaunch positioning leads with speed, not price.Notion · decisions · confirmed 12 Oct
CalendarLaunch target · Nov 18.Linear · NS-412
Northstarresearch memodocument·4 claims
Why teams switch
IllustrationThe market doesn’t need another feature. It needs a clearer reason to switch.
Seven of nine competitors lead with feature counts; none says how long the first order takes.[1]
Eleven of fourteen interviewed teams chose their current supplier for setup speed, then stayed out of habit.[2]
[1]competitor pages · 9[2]customer interviews · 14
- research
- Claude · research
- build
- Claude · implementation
- review
- GPT · independent review
- Mission · Brain · artifacts
- unchanged
The receipt names the model for each step and why Auto chose it. The reviewer is a model the builder did not use, when another is connected.
Act 02Modes
Say how hard it tries. Auto chooses who.
- FastQuick everyday work: the fastest connected model, low effort.
- AutoBoomship chooses the model and the effort for each task; planning and review escalate on their own.
- DeepMore exploration and verification: a frontier model at full effort.
- SwarmParallel approaches: the work is split into nodes that run side by side in their own worktrees, each verified, then merged.
Act 03The machine
Not a leaderboard. A machine you keep.
- Routing
- Auto picks a model for each step from the providers you connected: planning and review escalate on their own. The receipt says which model and why.
- Budgets
- Every run carries a token, cost and time budget, charged on each model turn. A reached limit stops the run, and the receipt says which one.
- Independence
- Change providers next month. The Brain, the projects, the connections, the schedules, the artifacts and the receipts stay where they are.
- Cost
- Inference is billed by your provider on your keys. Every call is counted in Usage by model, with its tokens and cost.
- Keys
- Encrypted with AES-256-GCM; only the last four characters are ever shown again. Where the deployment has managed keys, Auto can use them, and Settings says which providers are available.
- Privacy
- What reaches a model is the compiled context for the run, which the Context card shows you item by item, with a reason for each.
receipt · model
routingAuto · research → Gemini · build → Claude
cost$1.86
time48m 12s
budgetnot reached
IllustrationKeep your models. Upgrade everything around them.
Boomship Core$14 / monthfor everything around the model. Your provider bills the inference.