CCWAI VS Code Extension Status
Tracks completed and planned features for a VS Code extension that integrates AI context extraction and provider routing.
What this file does
Tracks completed and planned features for a VS Code extension that integrates AI context extraction and provider routing.
When to use it
- Planning the next development sprint for a VS Code extension
- Reviewing progress on a multi-package monorepo with CLI and extension
- Onboarding new contributors to understand extension capabilities and roadmap
Assumes this stack
CCWAI VS Code Extension Status
✅ Completed (v0.4.0 Phase 1 Foundation)
- Extension workspace scaffolding — Added dedicated
extension/package with manifest, build/publish scripts, TypeScript config, and dependency wiring so the VS Code bundle compiles alongsidecore/cli. - Activation + commands — Implemented
ccwai.analyzeContext,ccwai.planWithAI,ccwai.runCommand,ccwai.sendContext, andccwai.showThoughts, each invoking the sharedCCWAIEngine, streaming to the CCWAI output channel, and optionally persisting results. - Context extractor pipeline — Captures active editor, selection, workspace-relative path, and timestamp, trims content intelligently, and feeds formatted snapshots into both CCWAI and external providers.
- Context panel webview — Sidebar shows summary/snippet, git status (
git status --short --branchfallback togit diff --stat), and the three most recent extension thoughts, with buttons to trigger Analyze/Send actions via webview messaging. - Thoughts integration — Extension writes Markdown artifacts into
thoughts/shared/extension, exposes a quick-pick viewer, and keeps the panel in sync after each run. - Provider bridge v1 — Detects installed AI extensions by command IDs (Copilot/Codex/Gemini/DeepSeek), sends prompts directly when possible, or falls back to browser launch + clipboard hand-off; user preference set via
ccwai.preferredExternalProvider. - Provider routing indicators — Explicitly map known Copilot, Gemini, Codex, and DeepSeek command IDs, detect their availability, and surface their status inside the context panel so users know whether prompts go direct or via browser clipboard.
- Engine bridge — Safely discovers the CCWAI project root, validates
ccwai.config.json/commands, and instantiatesCCWAIEnginewith processcwdjuggling so CLI + extension share logic. - Workspace build —
npm run build --workspacesnow compiles core, CLI, and extension, ensuring a single command validates the full monorepo.
🚀 Next Steps (Prioritized TODOs)
- First-run / onboarding experience
- Add welcome panel or notification guiding users through configuring providers, running Analyze Context, and locating saved thoughts.
- Thoughts explorer view
- Replace the quick-pick list with a tree view showing shared vs personal thoughts; allow opening, deleting, or exporting directly from VS Code.
- Authentication / API key helpers
- Provide commands/UI to set OpenAI/Gemini/DeepSeek keys or launch OAuth/login flows for users without local keys.
- Command palette macros
- Add presets (e.g., “CCWAI: Analyze Current Git Diff”, “CCWAI: Draft PR Summary”) that prefill prompts using the context extractor + git summary.
- Release automation
- Create CI workflow to lint, build the extension, run smoke tests, and produce a
.vsix; document the publish process for the VS Code Marketplace.
- Create CI workflow to lint, build the extension, run smoke tests, and produce a
- Telemetry & diagnostics
- Optional (opt-in) logging of command usage/errors, surfaced via
ccwai doctorinside VS Code to aid troubleshooting.
- Optional (opt-in) logging of command usage/errors, surfaced via
- Documentation refresh
- Update
README.md/ docs with extension installation steps, screenshots, and the new settings (preferredExternalProvider,autoSaveThoughts,contextDepth).
- Update
We can now work through these in order—starting with provider routing polish and onboarding—while keeping the extension shippable at each increment.
What's inside
One status table with 9 completed items and a prioritized list of 7 next steps.
Change this for your project
- Replace
ccwaicommand IDs likeccwai.analyzeContextwith your own extension prefix - Replace
thoughts/shared/extensionpath with your own artifact storage location - Replace
ccwai.preferredExternalProvidersetting name with your own configuration key
Where it goes
Keep it in your repository where the agent or team that needs it will read it.
Worth borrowing
- Separating completed work from prioritized TODOs in a single document for clarity
- Using a provider bridge pattern to detect and route to multiple AI extensions
Related Documents
MCP Integration Workflows and Orchestration Guide
Defines enterprise-grade integration patterns, orchestration strategies, and advanced workflows for MCP servers with code examples.
AI-Generated Short-Form Video Platform
Defines a modular SaaS pipeline that transforms uploaded MP4 videos into polished short-form clips using AI agents and human review.
Valet V1 — Architecture & Implementation Plan
Defines a hosted background coding agent platform with isolated sandboxed dev environments, multiplayer sessions, and YAML-driven workflows.
GitHub Topics & Skills Monitor
Provides a structured GitHub monitoring strategy with search queries, watchlist, and notification options for AI skills, agents, and MCP servers.