Context-First Kanban for AI Agents - MCP server with web UI for managing AI agent workflows
npm install @sureshdsk/devflow-mcpKanban for AI Agents
DevFlow is a Kanban board where humans plan features and AI agents execute them. Agents "check in" to receive instructions and "check out" when done - creating a seamless handoff between sessions.
``bash`
npm install -g @sureshdsk/devflow-mcp
Requirements: Bun for the MCP server.
`bash`
devflow dev
Open http://localhost:3000 to see your Kanban board.
Claude Code CLI:
`bash`
claude mcp add devflow -- devflow mcp
Claude Desktop (~/Library/Application Support/Claude/claude_desktop_config.json):`json`
{
"mcpServers": {
"devflow": {
"command": "devflow",
"args": ["mcp"]
}
}
}
Cursor: Settings → Features → MCP → Add server with command devflow mcp
`
Create a project called "My App"
`
List all projects
`
Create a feature called "User Auth" in the My App project
`
Upload a planning doc to the User Auth feature:
$3
Create these tasks for the User Auth feature:
1. Set up JWT utilities (high priority)
2. Create login endpoint (high priority)
3. Add auth middleware (medium priority)
Show me all todo tasks
$3
Check me in to the "Set up JWT utilities" task
Log progress: Implemented token generation and validation
Check me out with summary: JWT utilities complete, added to src/lib/jwt.ts
$3
Mark the task as interrupted with context:
DONE: Token generation
REMAINING: Token validation
BLOCKED: Need Redis for refresh tokens
MCP Tools
| Category | Tools |
|----------|-------|
| Projects |
, create_project, get_project, get_or_create_project |
| Features | list_features, create_feature, get_feature |
| Tasks | list_tasks, create_task, create_tasks_bulk, get_task, update_task |
| Files | upload_file, list_files, get_file, update_file |
| Agent | check_in, check_out, log_activity, get_activity_log |Architecture
`
AI Agents ──► MCP Server ──► SQLite (~/.devflow/devflow.db)
│
▼
WebSocket ──► Web UI (localhost:3000)
Development
bash
git clone https://github.com/sureshdsk/devflow-mcp.git
cd devflow-mcp
bun install
bun dev # Web UI
bun run mcp # MCP server
``- Agent Guide - Detailed guide for AI agents
- Contributing - How to contribute
MIT