dflowcli
v1.0.4
Published
DevFlow Command Center CLI - Synchronized project execution control for AI-powered development
Downloads
18
Maintainers
Readme
dflowcli
DevFlow Command Center CLI - Synchronized project execution control for AI-powered development.
Installation
npm install -g dflowcliQuick Start
1. Initialize DevFlow in your project
cd your-project
devflow initThis will:
- Connect to your DevFlow Command Center project
- Generate a
CLAUDE.mdwith MCP tool documentation - Set up the
.devflow-context/directory for context sync - Configure Claude Code MCP integration
2. Monitor execution in real-time
devflow statusWatch tasks as they're assigned, in-progress, and completed.
3. Sync project context
devflow syncExports all project context to local markdown files:
.devflow-context/ACTIVE.md- Current task details.devflow-context/PROJECT.md- PRD, milestones, artifacts.devflow-context/CLIENT.md- Client info, communications.devflow-context/EXECUTION.md- All deliverables and tasks.devflow-context/DECISIONS.md- Decision log.devflow-context/ACTIVITY.md- Recent activity
Commands
| Command | Description |
|---------|-------------|
| devflow init | Initialize DevFlow in current project |
| devflow status | Show current execution status |
| devflow tasks | List all tasks and their status |
| devflow sync | Sync project context to local files |
| devflow watch | Watch for real-time execution updates |
| devflow login | Authenticate with DevFlow |
| devflow --help | Show all available commands |
Authentication
devflow login \
--url https://devflow.now \
--token YOUR_TOKEN \
--actor-id YOUR_USER_ID \
--actor-name "Your Name"Task Management
# Start a task
devflow task start <taskId>
# Block a task
devflow task block <taskId> --reason "Waiting for API credentials"
# Complete a task
devflow task complete <taskId> --note "Feature implemented and tested"
# Add a note to a task
devflow task note <taskId> --type decision --content "Using PostgreSQL for JSON support"
# List task notes
devflow task notes <taskId>Execution Monitoring
# View execution state
devflow exec status --project <PROJECT_ID>
# Watch real-time activity
devflow exec watch --project <PROJECT_ID>Environment Variables
Set these in your project's .env file:
DEVFLOW_PROJECT_ID=your-project-uuid
DEVFLOW_API_URL=https://devflow.now/apiClaude Code Integration
After running devflow init, your CLAUDE.md will contain instructions for Claude Code to use the MCP server. Install the MCP server:
npm install -g dflowmcpThe init command auto-configures your Claude Code MCP settings.
MCP Tools Available
| Tool | Description |
|------|-------------|
| get_active_task | Get currently assigned task |
| start_task | Start working on a task |
| complete_task | Mark task as done |
| block_task | Report a blocker |
| log_progress | Log progress updates |
| add_task_note | Add structured notes |
| sync_project_context | Sync context to local files |
Configuration
The CLI stores configuration in ~/.devflow/config.json:
{
"apiUrl": "https://devflow.now",
"token": "eyJhbGciOi...",
"actorId": "f47ac10b-58cc-4372-a567-0e02b2c3d479",
"actorName": "Sarah"
}Real-Time Synchronization
All changes are instantly visible across:
- Web UI (
devflow.now/command-center) - CLI (
devflow watch) - AI Agents (via MCP server)
This enables seamless human-agent collaboration.
Links
- Command Center: devflow.now/command-center
- Documentation: devflow.now/docs
- MCP Server: dflowmcp
- GitHub: github.com/devflow-app/devflow
License
MIT
