NWO Deer-Flow Integration - Tool #20 for NWO Conway Agents
Project description
NWO Deer-Flow Integration
Tool #20 for NWO Conway Agents - Advanced research and coding capabilities via ByteDance's Deer-Flow super agent harness.
Overview
NWO Deer-Flow integrates ByteDance's Deer-Flow - a powerful open-source super agent harness - into the NWO Robotics ecosystem as Tool #20 for Conway autonomous agents.
This integration enables NWO agents to:
- Conduct deep research tasks (minutes to hours)
- Generate code, reports, and presentations
- Create websites and visual content
- Spawn sub-agents for complex multi-step workflows
- Access sandboxed execution environments
Architecture
┌─────────────────────────────────────────────────────────────────┐
│ NWO CONWAY AGENT │
│ ┌─────────┐ ┌─────────┐ ┌─────────┐ ┌─────────────────────┐ │
│ │ Tool 1 │ │ Tool 2 │ │ ... │ │ Tool 20 │ │
│ │eml_regr │ │design_ │ │ │ │ deer_flow │ │
│ │ ess │ │ part │ │ │ │ (THIS PACKAGE) │ │
│ └────┬────┘ └────┬────┘ └─────────┘ └──────────┬──────────┘ │
│ └────────────┬─────────────────────────────┘ │
│ │ │
│ NWO Agent Runner (Cloudflare Worker) │
│ │ │
└────────────────────┼────────────────────────────────────────────┘
│
▼ HTTP API
┌──────────────────────────────┐
│ NWO Deer-Flow Service │
│ (Self-hosted or NWO-hosted) │
└──────────────┬───────────────┘
│
┌──────────────┼──────────────┐
▼ ▼ ▼
┌─────────┐ ┌─────────┐ ┌─────────────┐
│ LLM │ │ Sandbox │ │ Memory │
│ Provider│ │Execution│ │ Store │
└─────────┘ └─────────┘ └─────────────┘
Installation
For End Users (Connect Your Own LLM)
pip install nwo-deerflow
Set your environment variables:
export OPENAI_API_KEY="sk-..."
# OR
export ANTHROPIC_API_KEY="sk-ant-..."
# OR
export MOONSHOT_API_KEY="sk-..."
Run locally:
nwo-deerflow serve
For NWO Conway Agent Integration
The NWO Deer-Flow service is automatically available to all Conway agents as Tool #20 when deployed to the NWO infrastructure.
Quick Start
1. Local Development Mode
from nwo_deerflow import DeerFlowClient
# Initialize client
client = DeerFlowClient(
api_base="http://localhost:8001",
api_key="your-api-key" # Optional for local
)
# Submit a research task
response = client.submit_task(
prompt="Research the latest developments in humanoid robotics and create a summary report",
mode="pro", # flash, standard, pro, ultra
thread_id="my-thread-001"
)
print(f"Task ID: {response['task_id']}")
print(f"Status: {response['status']}")
2. Asynchronous Task Execution
import asyncio
from nwo_deerflow import DeerFlowClient
async def run_research():
client = DeerFlowClient()
# Submit and wait for completion
result = await client.submit_and_wait(
prompt="Create a Python script that analyzes crypto market trends",
mode="ultra", # Uses sub-agents for complex coding
timeout=3600 # Wait up to 1 hour
)
print(f"Result: {result['output']}")
print(f"Artifacts: {result['artifacts']}")
asyncio.run(run_research())
3. Conway Agent Tool Usage
When used as Tool #20 in a Conway agent:
{
"type": "deer_flow",
"args": {
"prompt": "Research quantum computing applications in robotics and generate a slide deck",
"mode": "pro",
"output_format": "slides",
"max_duration_minutes": 30
},
"note": "Deep research on quantum robotics"
}
Deployment Options
Option A: Self-Hosted (Bring Your Own API Keys)
Deploy on your own infrastructure with your own LLM API keys:
# Clone the repository
git clone https://github.com/RedCiprianPater/nwo-deerflow.git
cd nwo-deerflow
# Configure
make setup
# Deploy with Docker
make docker-start
Benefits:
- Full control over data and execution
- Use your own LLM API keys
- Custom skill configurations
- Private sandbox environments
Option B: NWO-Hosted (Included in Conway Agent)
Deploy as part of the NWO Conway agent infrastructure:
# Deployed automatically with NWO Agent Runner
# No additional configuration needed
# Usage billed through Conway agent's operational balance
Benefits:
- Zero setup required
- Integrated with NWO agent ecosystem
- Automatic scaling
- Shared memory across NWO tools
Cloudflare Worker Integration
The NWO Deer-Flow integration includes a Cloudflare Worker that acts as a lightweight gateway between Conway agents and the Deer-Flow service:
// worker.js - Deploy to Cloudflare Workers
import { DeerFlowTool } from './tools/deerflow';
export default {
async fetch(request, env, ctx) {
const url = new URL(request.url);
if (url.pathname === '/api/tool/20/deer-flow') {
const tool = new DeerFlowTool(env);
return await tool.handle(request);
}
return new Response('Not Found', { status: 404 });
}
};
Configuration
Environment Variables
| Variable | Description | Default |
|---|---|---|
DEERFLOW_API_BASE |
Deer-Flow API endpoint | http://localhost:8001 |
DEERFLOW_API_KEY |
API key for authentication | None |
OPENAI_API_KEY |
OpenAI API key | None |
ANTHROPIC_API_KEY |
Anthropic API key | None |
MOONSHOT_API_KEY |
Moonshot/Kimi API key | None |
DEERFLOW_SANDBOX_MODE |
Sandbox execution mode | docker |
DEERFLOW_MAX_DURATION |
Max task duration (minutes) | 60 |
config.yaml
# Deer-Flow Configuration
models:
- name: gpt-4o
display_name: GPT-4o
use: langchain_openai:ChatOpenAI
model: gpt-4o
api_key: $OPENAI_API_KEY
- name: kimi-k2.5
display_name: Kimi K2.5
use: langchain_openai:ChatOpenAI
model: kimi-k2.5
api_key: $MOONSHOT_API_KEY
base_url: https://api.moonshot.ai/v1
sandbox:
use: deerflow.community.aio_sandbox:AioSandboxProvider
# Options: local, docker, kubernetes
skills:
- research
- report-generation
- slide-creation
- web-page
- image-generation
- code-generation
API Reference
Submit Task
POST /api/tasks
Content-Type: application/json
{
"prompt": "Research topic and generate output",
"mode": "pro",
"thread_id": "optional-thread-id",
"context": {
"previous_findings": [...],
"constraints": [...]
}
}
Get Task Status
GET /api/tasks/{task_id}
List Tasks
GET /api/tasks?status=running&limit=10
Cancel Task
POST /api/tasks/{task_id}/cancel
Integration with NWO Conway Relayer
The Deer-Flow tool integrates seamlessly with the NWO Conway Relayer system:
# Conway agent using Deer-Flow via relayer
from nwo_conway_relayer import ConwayAgent
agent = ConwayAgent(
wallet_address="0x...",
relayer_url="https://nwo-conway-relayer.onrender.com"
)
# Tool 20: Deer-Flow research
result = agent.execute_tool(
tool_id=20,
params={
"prompt": "Analyze DeFi yield farming strategies",
"mode": "ultra",
"output_format": "report"
}
)
Modes Explained
| Mode | Description | Duration | Use Case |
|---|---|---|---|
flash |
Quick answers | 1-5 min | Simple queries |
standard |
Balanced research | 5-15 min | General research |
pro |
Deep research with planning | 15-45 min | Complex analysis |
ultra |
Multi-agent orchestration | 30-120 min | Large projects |
Skills Available
Deer-Flow comes with built-in skills that NWO agents can leverage:
- Research - Deep web research with source verification
- Report Generation - Create structured reports
- Slide Creation - Generate presentation decks
- Web Page - Create and deploy websites
- Image Generation - Generate visuals and diagrams
- Code Generation - Write and test code
Development
# Clone repository
git clone https://github.com/RedCiprianPater/nwo-deerflow.git
cd nwo-deerflow
# Install dependencies
make install
# Run tests
make test
# Start development server
make dev
Contributing
We welcome contributions! Please see CONTRIBUTING.md for guidelines.
License
MIT License - see LICENSE file.
Acknowledgments
- ByteDance Deer-Flow - The underlying super agent harness
- NWO Robotics - Conway agent ecosystem
- LangChain - LLM framework
- LangGraph - Agent orchestration
Support
- GitHub Issues: github.com/RedCiprianPater/nwo-deerflow/issues
- NWO Discord: discord.gg/nwo-robotics
Built with ❤️ by the NWO Robotics team
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file nwo_deerflow-0.1.0.tar.gz.
File metadata
- Download URL: nwo_deerflow-0.1.0.tar.gz
- Upload date:
- Size: 14.1 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.13.12
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
98e1a691dc85fc4432ec6b3237280e6fbc63d3fb73409b1ead27fe694275fa28
|
|
| MD5 |
cc6688e8a6cae049f1c2c513e9ae46fc
|
|
| BLAKE2b-256 |
e4c39e479bb3eecdcd7b1d970bd5d3e2c5694457706534280cbedd0afbf3d951
|
File details
Details for the file nwo_deerflow-0.1.0-py3-none-any.whl.
File metadata
- Download URL: nwo_deerflow-0.1.0-py3-none-any.whl
- Upload date:
- Size: 16.1 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.1.0 CPython/3.13.12
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
232fb9716cae3cc76b82adfc6cb1e17c69834fd58464929a6199f5f409bc60c2
|
|
| MD5 |
89040c5505eb74f5b351376322b2f34d
|
|
| BLAKE2b-256 |
58648e220d80958ac68e376b7edae7ec9633d9a02ec9a1d2a45ca0389926a497
|