🚀 FlowMind
A Lightweight Multi-Agent Automation Platform for Enterprise Tasks
FlowMind is a native Python automation framework that makes building workflows simple without the complexity of LangChain, Airflow, or enterprise tools.
✨ Features
- 🎯 Zero Complexity - Simple task-based API, no abstractions
- ⚡ Zero Dependencies - Core package works offline, no external deps
- 🔌 Plugin Architecture - Extend with custom tasks
- 📦 Built-in Tasks - 7+ ready-to-use task types
- 🔄 Smart Context - Tasks share data seamlessly
- ⏰ Scheduler - Cron-like periodic execution
- 🐍 Pure Python - No Docker, K8s, or containers needed
🎯 Why FlowMind?
| Feature | FlowMind | LangChain | Airflow | n8n |
|---|---|---|---|---|
| Simplicity | ⭐⭐⭐⭐⭐ | ⭐⭐ | ⭐⭐ | ⭐⭐⭐⭐ |
| Offline | ✅ | ❌ | ✅ | ❌ |
| Pure Python | ✅ | ⚠️ | ⚠️ | ❌ |
| No Dependencies | ✅ (core) | ❌ | ❌ | ❌ |
| Learning Curve | 5 min | 2 days | 1 week | 1 day |
📦 Installation
# Install core (zero dependencies)
pip install flowmind
# Install with optional features
pip install flowmind[full]
🚀 Quick Start
Example 1: Simple Data Pipeline
from flowmind import FlowAgent
# Create an agent
agent = FlowAgent("data_processor")
# Add tasks
agent.add_task("file", operation="read", file_path="data.json", name="load")
agent.add_task("data", operation="filter", input="${load.output}", condition="price > 100", name="filter")
agent.add_task("file", operation="write", file_path="results.json", content="${filter.output}", as_json=True)
# Run workflow
results = agent.run()
print(f"✅ Processed {len(results)} tasks")
Example 2: Web Scraping & Email
from flowmind import FlowAgent
agent = FlowAgent("price_monitor")
# Scrape website
agent.add_task("web", operation="get", url="https://api.example.com/price", name="fetch")
# Check condition
agent.add_task("data", operation="transform", input="${fetch.content}", name="check")
# Send email if price dropped
agent.add_task("email",
operation="send",
to="admin@example.com",
subject="Price Alert!",
body="Price dropped to ${fetch.content.price}",
if_condition="${fetch.content.price} < 100",
name="notify"
)
# Schedule to run every hour
agent.schedule(every="1h")
agent.start()
Example 3: ML Pipeline
from flowmind import FlowAgent
agent = FlowAgent("ml_pipeline")
# Load data
agent.add_task("file", operation="read", file_path="data.csv", name="load")
# Train model
agent.add_task("ml",
operation="train",
model="random_forest",
data="${load.output}",
target="price",
name="train"
)
# Make predictions
agent.add_task("ml",
operation="predict",
model="${train.model}",
data="test.csv",
name="predict"
)
# Save results
agent.add_task("file",
operation="write",
file_path="predictions.json",
content="${predict.output}",
as_json=True
)
agent.run()
📚 Built-in Tasks
1. FileTask - File Operations
agent.add_task("file", operation="read", file_path="data.json")
agent.add_task("file", operation="write", file_path="output.txt", content="Hello")
agent.add_task("file", operation="copy", source="a.txt", destination="b.txt")
agent.add_task("file", operation="delete", file_path="temp.txt")
agent.add_task("file", operation="list", directory=".", pattern="*.py")
2. WebTask - HTTP & Scraping
agent.add_task("web", operation="get", url="https://api.example.com/data")
agent.add_task("web", operation="post", url="https://api.example.com", data={"key": "value"})
agent.add_task("web", operation="download", url="https://example.com/file.pdf", output="file.pdf")
3. DataTask - Data Transformation
agent.add_task("data", operation="filter", input="${load.output}", condition="price > 100")
agent.add_task("data", operation="aggregate", input="${data}", agg_type="sum", field="amount")
agent.add_task("data", operation="sort", input="${data}", key="price", reverse=True)
4. EmailTask - Email Operations
agent.add_task("email", operation="send", to="user@example.com", subject="Alert", body="Message")
agent.add_task("email", operation="classify", content="${email_text}", categories=["urgent", "spam", "normal"])
5. PDFTask - PDF Extraction
agent.add_task("pdf", operation="extract", file_path="invoice.pdf", extract=["text", "tables"])
6. MLTask - Machine Learning
agent.add_task("ml", operation="train", model="random_forest", target="price")
agent.add_task("ml", operation="predict", model="${train.model}", data="test.csv")
7. ShellTask - Shell Commands
agent.add_task("shell", command="ls -la", name="list")
agent.add_task("shell", command="python script.py", timeout=60)
🔌 Plugin System
Create custom tasks:
from flowmind import BaseTask, TaskResult, TaskStatus, register_task
@register_task("my_custom_task")
class MyCustomTask(BaseTask):
def execute(self, context):
# Your logic here
result = do_something()
return TaskResult(
status=TaskStatus.SUCCESS,
output=result
)
# Use it
agent.add_task("my_custom_task", name="custom", param1="value1")
🎯 Variable Substitution
Tasks can reference each other's outputs:
agent.add_task("file", operation="read", file_path="data.json", name="load")
agent.add_task("data", operation="filter", input="${load.output}", condition="price > 100")
agent.add_task("file", operation="write", content="${filter.output}", file_path="result.json")
⏰ Scheduling
Run workflows periodically:
agent = FlowAgent("scheduled_job")
agent.add_task("shell", command="python backup.py")
agent.schedule(every="1h") # Run every hour
agent.start()
# Supported intervals: "30s", "5m", "1h", "1d"
🎨 Conditional Execution
Run tasks based on conditions:
agent.add_task("web", operation="get", url="https://api.example.com/status", name="check")
agent.add_task("email",
operation="send",
to="admin@example.com",
subject="System Down!",
if_condition="${check.status_code} != 200", # Only run if check failed
name="alert"
)
📖 Documentation
FlowAgent API
agent = FlowAgent(name="workflow", verbose=True)
agent.add_task(task_type, name=None, **config)
agent.run(stop_on_error=True)
agent.schedule(every="5m")
agent.start()
agent.stop()
agent.get_result(task_name)
agent.clear()
Task Configuration
Every task supports:
name: Optional task name (auto-generated if not provided)if_condition: Conditional execution (e.g.,"${prev.status} == 200")- Task-specific parameters (see Built-in Tasks section)
🌟 Use Cases
- Data Pipelines - ETL, data cleaning, transformation
- Web Automation - Scraping, API integration, monitoring
- File Processing - Batch processing, format conversion
- Email Automation - Alerts, notifications, classification
- ML Workflows - Train, predict, deploy models
- DevOps Tasks - Deployment, monitoring, backups
- Business Automation - Invoice processing, reporting
🎯 Target Audience
- Python Developers - Need simple automation
- Data Scientists - Build ML pipelines
- DevOps Engineers - Automate workflows
- Small Businesses - No budget for enterprise tools
- Students - Learn automation concepts
🚀 Why Not LangChain?
LangChain is powerful but complex:
- Steep learning curve
- Heavy dependencies
- LLM-focused (not general automation)
- Requires cloud services
FlowMind is:
- Simple & intuitive
- Zero dependencies (core)
- General-purpose automation
- Works offline
📊 Performance
- Lightweight: ~50KB core package
- Fast: Minimal overhead, pure Python
- Efficient: No unnecessary abstractions
🤝 Contributing
Contributions welcome! Please check issues or submit PRs.
git clone https://github.com/idrissbado/flowmind.git
cd flowmind
pip install -e ".[full]"
📄 License
MIT License - see LICENSE file for details.
🙏 Acknowledgments
Built with ❤️ for developers who value simplicity.
🔗 Links
- GitHub: https://github.com/idrissbado/flowmind
- PyPI: https://pypi.org/project/flowmind/
- Documentation: https://github.com/idrissbado/flowmind#readme
- Issues: https://github.com/idrissbado/flowmind/issues
📧 Contact
Idriss Bado
- GitHub: @idrissbado
- Email: idrissbado@gmail.com
Made with 🚀 by Idriss Bado
"From Complex AI Agents to Simple Python Workflows"
Metadata
Release files for flowmind 0.1.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| flowmind-0.1.0.tar.gz | 29.0 kB | Without provenance | Details for flowmind-0.1.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| flowmind-0.1.0-py3-none-any.whl | Python 3 | none | any | Details for flowmind-0.1.0-py3-none-any.whl |
Total release size: 60.4 kB
Release files / flowmind-0.1.0.tar.gz
| Download URL | flowmind-0.1.0.tar.gz |
|---|---|
| Size | 29.0 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
84454719f71e1720b119d80479923f17b9b307bd032bac3e1f48490322509576
|
|
BLAKE2b-256 checksum How to use checksums |
806ea12acdbaadcd225320d940d9779dcbb2f09c3d381d611ba7eb5b2873c969
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.12.0
|
Release files / flowmind-0.1.0-py3-none-any.whl
| Download URL | flowmind-0.1.0-py3-none-any.whl |
|---|---|
| Size | 31.5 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
db2659156025bf6ca32647d6a32dee594233cbde8cbf90dab23a566ab403fa0e
|
|
BLAKE2b-256 checksum How to use checksums |
4241c0a591808cd781afb76db6f2776118bcffc5477e65f3192d6db78a76896f
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.12.0
|
Security for flowmind 0.1.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.