Skip to content

PRODUCT_VISION.md

Latest commit

 

History

History
122 lines (97 loc) · 7.18 KB

PRODUCT_VISION.md

File metadata and controls

122 lines (97 loc) · 7.18 KB

TaskWing: Local-First AI Knowledge Layer

TaskWing extracts architectural knowledge from codebases and stores it locally, giving every AI coding tool instant context without your knowledge base leaving your machine.

Vision Statement

TaskWing is the local-first knowledge layer for AI-assisted development.

We extract your architecture -- decisions, patterns, constraints -- into a local SQLite database. Every AI tool gets instant context via MCP, without your knowledge base leaving your machine.

Private by Architecture

Property How
Knowledge stored locally SQLite on your filesystem. No cloud database, no sync.
No account required Install via brew and use immediately. Zero data collection surface.
AI tools connect locally MCP queries served over local stdio. No network calls for queries.
Air-gappable Ollama support for fully offline operation. Zero external dependencies.
Open source (MIT) Audit every line. Fork it. Run it on your own terms.
No vendor kill switch MIT license, SQLite storage, standard MCP protocol.

What TaskWing controls: During bootstrap, code context is processed by your chosen LLM provider (cloud or Ollama for full local). After extraction, your knowledge base is stored and queried locally. MCP responses are served over local stdio and never touch the network.

What your AI tool controls: Cloud-based AI tools (Claude, Cursor, Copilot) send conversations -- including TaskWing's MCP responses -- to their own servers per their privacy policies. TaskWing cannot control this. To keep everything local, use Ollama for bootstrap and a local AI tool for queries.

Ecosystem Support

Supported Models

OpenAI Anthropic Google Gemini AWS Bedrock Ollama

Works With

Claude Code OpenAI Codex Cursor GitHub Copilot Gemini CLI OpenCode

Brand names and logos are trademarks of their respective owners; usage here indicates compatibility, not endorsement.

Architecture

┌─────────────────────────────────────────────────────────┐
│                    USER INTERFACE                        │
│  /taskwing:plan       │  /taskwing:next  │  /taskwing:done   │
└─────────────────────────────────────────────────────────┘
                              │
                              v
┌─────────────────────────────────────────────────────────┐
│                   TASK GENERATION                        │
│  Analyze goal -> Query knowledge graph -> Generate tasks │
└─────────────────────────────────────────────────────────┘
                              │
                              v
┌─────────────────────────────────────────────────────────┐
│            LOCAL KNOWLEDGE GRAPH (The Moat)              │
│  Features │ Patterns │ Decisions │ Constraints │ Files  │
│            Stored in local SQLite -- never synced        │
└─────────────────────────────────────────────────────────┘
                              │
                              v
┌─────────────────────────────────────────────────────────┐
│                 LOCAL MCP SERVER (stdio)                  │
│  Claude │ Cursor │ Copilot │ Codex -- all get context   │
│            No network calls for queries                  │
└─────────────────────────────────────────────────────────┘

Core Commands

  • taskwing bootstrap
  • taskwing ask "<query>"
  • taskwing task
  • taskwing mcp
  • taskwing doctor
  • taskwing config
  • taskwing start

MCP Tools

Tool Description
ask Search project knowledge (decisions, patterns, constraints)
task Unified task lifecycle (next, current, start, complete)
plan Plan management (clarify, decompose, expand, generate, finalize, audit)
code Code intelligence (find, search, explain, callers, impact, simplify)
debug Diagnose issues systematically with AI-powered analysis
remember Store knowledge in project memory

Success Metrics

  1. Task accuracy: generated tasks reference correct files and patterns.
  2. Developer adoption: daily active users running /taskwing:plan.
  3. Context utilization: MCP queries per plan execution.
  4. Time-to-root-cause: bug investigations with TaskWing context vs. without.

Monetization (Future)

Tier Price Features
Open Source Free Full CLI, local knowledge graph
Team $29/seat/mo Shared knowledge graph, team sync
Enterprise Custom SSO, audit, on-prem

The local knowledge graph is the moat. AI-assisted development is the product.