0.2.7

AgentEazy

Turn any Python utility into a callable AI tool in one command.

Your code becomes a live API that any AI agent can find, call, and pay — without rewriting a single line.

Website · Dashboard · Registry · Docs

---

Try it right now

No install needed. Call a live agent from your terminal:

curl -s -X POST https://simondusable--agenteazy-gateway-serve.modal.run/agent/zxcvbn-python/ \
  -H "Content-Type: application/json" \
  -d '{"verb":"DO","payload":{"data":{"password":"monkey123"}}}' | python3 -m json.tool

{
  "status": "completed",
  "output": {
    "score": 1,
    "feedback": {
      "warning": "This is a very common password.",
      "suggestions": ["Add another word or two. Uncommon words are better."]
    }
  }
}

That's a real GitHub repo (zxcvbn-python), auto-wrapped and live. 18+ more in the registry.

---

Deploy your own agent

pip install agenteazy
agenteazy signup your-github-username --email you@example.com
agenteazy deploy github.com/you/your-repo

Done. Your repo is live. Every AI agent on earth can now find and call it.

What happens under the hood

1. Analyze — Clones the repo, parses the AST, detects the main public API function
2. Wrap — Generates a FastAPI wrapper that handles imports, dependencies, and dispatch
3. Upload — Pushes to serverless infrastructure (idle agents cost $0)
4. Register — Adds to the public registry so other agents can discover it

---

Python SDK

from agenteazy import AgentEazy

client = AgentEazy()

# Discover agents
agents = client.find("password strength")

# Call agents
result = client.do("zxcvbn-python", {"password": "monkey123"})
# → {"score": 1, "feedback": {"warning": "This is a very common password."}}

result = client.do("dateparser", {"date_string": "tomorrow at 3pm"})
# → "2026-03-17T15:00:00"

result = client.do("langdetect", {"text": "Bonjour le monde"})
# → "fr"

result = client.do("python-ftfy", {"text": "émile dupont"})
# → "émile dupont"

result = client.do("autopep8", {"source": "x=   1\nif  x==1:\n  print( 'hello' )"})
# → "x = 1\nif x == 1:\n    print('hello')\n"

---

LangChain Integration

Every agent becomes a LangChain BaseTool:

pip install agenteazy[langchain]

from agenteazy.integrations.langchain import AgentEazyTool, AgentEazyToolkit

tool = AgentEazyTool.from_agent("zxcvbn-python")
result = tool.invoke({"password": "test123"})

# Discover multiple tools from the registry
toolkit = AgentEazyToolkit()
tools = toolkit.get_tools(query="text processing", limit=5)

---

CrewAI Integration

pip install agenteazy[crewai]

from agenteazy.integrations.crewai import AgentEazyCrewTool

tool = AgentEazyCrewTool.from_agent("zxcvbn-python")

---

Live Agent Registry

18 agents you can call right now, all auto-wrapped from real GitHub repos:

Browse all at agenteazy.com/agents.

---

What repos work best

AgentEazy auto-detects the right entry point for stateless Python utilities — functions that take input and return output.

Override the entry point when needed:

agenteazy deploy github.com/you/repo --entry "mypackage/core.py:process"
agenteazy deploy github.com/you/repo --entry "mypackage/model.py:MyClass.predict"

---

TollBooth — Agent Credits

Agents can charge credits per call. Set a price and the gateway handles billing.

agenteazy deploy github.com/you/your-repo --price 10

• Free by default — no credits needed to call free agents
• You set the price — 1 credit, 10, 100 — your choice
• 80/20 split — you keep 80%, platform takes 20% for infrastructure
• 50 free credits on signup, buy more at agenteazy.com/dashboard

---

AgentLang — 10 Universal Verbs

One protocol for all agents. One HTTP POST, one envelope:

POST /agent/{name}/
{"verb": "DO", "payload": {"data": {"input": "hello"}}}

---

CLI Reference

# Deploy
agenteazy deploy <repo>                    # Full pipeline
agenteazy deploy <repo> --local            # Test locally first
agenteazy deploy <repo> --price 10         # Paid agent
agenteazy deploy <repo> --entry "f.py:fn"  # Override entry point
agenteazy deploy <repo> --env KEY=VALUE    # Inject env vars

# Analyze & Wrap
agenteazy analyze <repo>                   # Inspect without deploying
agenteazy wrap <repo>                      # Generate wrapper only
agenteazy batch-deploy <repos.txt>         # Deploy from file

# Account
agenteazy signup <username> --email <email>  # Get API key + 50 credits
agenteazy balance                            # Check credits
agenteazy transactions                       # View history

# Registry
agenteazy search "query"                   # Find agents
agenteazy list                             # List all agents
agenteazy status                           # Health check

---

Architecture

agenteazy deploy <repo>
    │
    ├── ANALYZE  → Clone, parse AST, detect public API function
    ├── WRAP     → Generate FastAPI wrapper with full import handling
    ├── UPLOAD   → Push to serverless volume (idle = $0)
    ├── REGISTER → Add to searchable registry
    └── LIVE     → Callable at gateway/agent/{name}/

Gateway (serves all agents)  ←→  Registry (search + credits)
    │                                    │
    └── Agent Volume                     └── TollBooth
        /agents/zxcvbn-python/               Signup, balance,
        /agents/langdetect/                  deduct, earn, transfer
        /agents/dateparser/
        ... (18 agents)

---

Contributing

Best first contribution: OpenAI Function Calling integration. Same pattern as the LangChain wrapper — one file, self-contained. See agenteazy/integrations/langchain.py as a template.

git clone https://github.com/SimonSliman/agenteazy
cd agenteazy
pip install -e ".[dev]"
pytest tests/ -q  # 20 tests

---

License

MIT