v0.1.0-alpha138 tests. 58 skills. 4 provider lanes. One compounding system.
AIvan is a private developer system with public proof. Evidence-first architecture, compounding memory, and multi-provider intelligence — shipped, not promised.
Proof strip
The fastest way to trust AIvan is to see the visible command lanes already working in the repo.
Natural-language transforms
live lanenode ./bin/dev.js transform ./examples/orders.json "Sort by total descending and keep top 10"Intent becomes executable local data work instead of a one-off answer.
Structured-data transforms already run today for JSON, JSONL, and NDJSON inputs.Provider routing
live lanenode ./bin/dev.js transform ./examples/events.jsonl "Group by region" --provider claudeProvider choice is a runtime lane, not a redesign project.
The same task can move across Gemini, Claude, OpenAI, and local OpenAI-compatible endpoints.Workflow lane
live lanenode ./bin/dev.js workflow --helpWorkflow automation is already a first-class command surface.
The direction stays DAG-based task execution rather than a pile of prompt snippets.System architecture
Surfaces, core, knowledge, and delivery — each layer is shipped and working, not theoretical.
Four layers that make AIvan a system instead of a script.
Signals
These are not projections. Every metric comes from the current state of the AIvan repo and orchestration layer.
Coverage spans runtime, providers, scaffold, visualizer, and workflow lanes.
Planning, execution, verification, and recovery are already shaped by reusable skills.
The system is already wired for live external context instead of isolated prompt work.
Configured in AIvan settings as part of the working orchestrator surface.
Routing already spans explore, plan, bash, review, feature, and execution roles.
Hooks keep learning, status, and cleanup moving without manual babysitting.
Access tracks
AIvan is selective by design. Pick the track that matches your intent.
For people who want to test flows, pressure-test positioning, or help shape the product surface.
Visible product context, curated walkthroughs, and direct feedback loops into the roadmap.
For teams evaluating whether AIvan's transform, workflow, or orchestration model matches a real operating problem.
A focused conversation around use case fit, boundaries, and what a safe first integration would look like.
For senior builders who want to challenge the architecture, execution model, or product claims.
Concrete proof surfaces, architecture mapping, and discussion anchored in shipped behavior instead of hype.