CodeAI Hub is a Visual Studio Code extension + standalone Project Manager (CEF) that unifies multiple AI providers behind a single, type-safe orchestration layer.
- SolidWorks-WorkFlow docs index:
doc/SolidWorks-WorkFlow/Docs_Index.md - System SSOT:
doc/SolidWorks-WorkFlow/System/SystemArchitecture.md - Session input lock SSOT:
doc/SolidWorks-WorkFlow/Contracts/SessionInputLock_SSOT_StateMachine.md - Bug registry:
doc/BugRegistry.md
- Universal provider usage limits:
Claude,Codex, andGemininow use a shared module inpackages/core, where the live provider surface is the primary source and provider-specific fallbacks remain secondary. - Codex live-source migration: usage limits are read through
runtime payload -> app-server account/rateLimits/read -> rollout JSONL fallback, without a mandatory dependency on PTY/TUI/status. - WebSocket replay hardening: stateful
usage_limitssignals are replayed after websocket connect and after workspace scope changes, soCodexlimits are not lost inProject Manager/Session UIduring late attach or rebind. - Gemini dialog segmentation:
Gemini_Modulenow flushes assistant segments on everyfinishedevent and no longer duplicates them with a final aggregate block when segmented replies were already emitted through dialog history. - Flow-node continuity turn boundary: threshold-trigger continuity for document/workflow nodes can no longer interrupt an active user one-shot turn on the first
token_usage; rollover is allowed only afterturn_completedor by trailing usage in pending post-turn arbitration. - Release validation on March 15, 2026: a manual smoke test for the
Geminidocument node confirmed that inv1.1.730the active one-shot turn finishes completely before continuity handoff/bootstrap begins. - Session UI hardening: usage limits are now cached by the canonical
providerScopeKey, and theSession ID barrenders provider-aware labels from the shared snapshot instead of the old hardcodedsession/weekly. - Diagnostics: the shared usage-limits facade and
Codexintegration now expose source-aware diagnostics forcache_hit,fresh_read,fallback_cached, andunavailable, making refresh/fallback behavior easier to investigate. - Description cleanup baseline: the product remains on the canonical
questionnaire.md->Final_Description.mdflow without legacy↻ Restart attemptsemantics and without the olddescription.mdlabel in the tree/main area. - Core/runtime baseline: active artifact persistence remains on the canonical
/api/v1/orchestrator/artifact-upsert; obsolete restart-era transport does not return to the stable line. - Documentation governance: before
doc/TODO/todo-plan.md, every new scope must first live indoc/SolidWorks-WorkFlow/Plans/, while implemented SSOT remains only inSystem/,Clusters/,Modules/, andContracts/. - Agent instructions:
AGENTS.mdis the only git-tracked source of instructions; localGEMINI.mdand.claude/CLAUDE.mdare reduced to short redirect files. - Release pipeline: local
build-allmust raise the unified version to1.1.730and rebuild provider/core/ui/launcher artifacts on top of that baseline.
Previous releases (summary): the 1.1.57x–1.1.719 series focused on SSOT routing (dialog vs runtime), snapshot-first lock/usage authority, continuity/resume reliability across providers, Virtual Simulation workflow, Diagram Modules / Facades workflow, workflow handoff UX, panel sync in Project Manager, and later PM hydration/workflow-state experiments that are intentionally not part of this stable baseline release.
- Unified provider orchestration: launch Claude, Codex, or Gemini sessions from an identical picker; the dialog surfaces connection state, enforces one-provider selection, and reminds you to install/authenticate matching CLIs.
- Idea Collector flow: Codex and Claude sessions can launch a guided idea collection flow that produces structured Idea.md drafts.
- Persistent standalone UI: the macOS launcher (CEF) stores window position and size in real time, so Project Manager reopens exactly where you left it—even across monitor changes.
- Offline-first packaging: manifests point to the local
~/.codeai-hub/releases/cache, and build scripts publish fresh tarballs for core, launcher, and provider modules without relying on GitHub downloads. - Quality guardrails: Ultracite architecture rules, jscpd duplication scans, ts-prune export checks, and Biome formatting are orchestrated through Husky pre-commit/pre-push hooks.
git clone https://github.com/OleynikAleksandr/CodeAI-Hub.git
cd CodeAI-Hub
npm installBefore starting, read doc/SolidWorks-WorkFlow/Docs_Index.md and follow the SSOT contracts in doc/SolidWorks-WorkFlow/Contracts/ (especially Contracts/Workflow_CLI.md) to configure provider CLIs and SDKs.
- Install dependencies
npm install npm run setup:hooks # installs Husky git hooks - Implement changes in
src/andpackages/**(micro-classes + facades; keep files under 300 lines). - Run quality checks before committing:
npm run quality # architecture gate + Ultracite lint npm run check:tsprune # detect unused exports npm run compile # ensure TypeScript builds cleanly
./scripts/build-all.sh
./scripts/build-release.sh --use-current-versionmedia/ Bundled webview assets (CSS + JS) shipped with the extension.
media/react-chat.js React bundle generated by the webview build script.
src/core/webview-module/ HTML scaffold that injects the webview assets.
src/extension-module/ Extension host micro-classes.
src/extension.ts Entry point registering the webview provider.
scripts/ Quality and release automation.
doc/ Architecture and knowledge base.
License information will be added in a future update. Until then, treat the repository as proprietary and request permission before redistribution.