Agent Builder Studio
One app for the agent design → run → package lifecycle — a visual canvas authors and runs a governed spec; a deterministic engine packages it. Local-first.
The Problem
Designing an agent, running it against a real model, and packaging it for reuse are usually three disconnected tools — a whiteboard, a script, and a build step — with the truth scattered across all three. The graph you drew drifts from the code you ran, which drifts from the thing you shipped.
What I Built
Agent Builder Studio is one app for the whole design → run → package lifecycle. A visual canvas (Studio) authors a governed agent spec and runs it live with per-node streaming; a deterministic packaging engine (Builder) turns that spec into a shippable package. Local-first by default everywhere — MLX-first model execution with an Ollama fallback and key-gated cloud.
It consolidates two earlier projects — Agent Builder (the evaluation/packaging workbench) and Agent Studio (the runtime canvas) — plus a standalone Chief-of-Staff product, merged via git subtree so the history is preserved.
Architecture Principle
The agent spec is the single source of truth; the canvas layout, run transcripts, and the generated package are all derived, rebuildable projections. A derived layer never becomes the source of truth. (Adopted from build-loop-memory’s AI-first data architecture.)
Monorepo
npm install links all workspaces and npm test runs the deterministic suite across the shared agent-spec + local-llm packages and every app. Per-app build/deploy stays independent.
Status
Mid-merge — the two source apps are being collapsed into a single apps/studio, with phased compile + test gates between steps.