sagewai 0.1.1

Open-source AI agent platform for production workloads — full-stack, durable, safe, and model-agnostic.

Sagewai

Agent Infrastructure You Own

The open-source platform for AI agents that run, remember, and report. Build agents in Python. Operate them from a dashboard. Own everything.

The 5 Pillars

Pillar	Description
SDK	Build agents with multi-model support, tools, memory, guardrails, and durable workflows
Registry	Store, version, discover, and govern AI agents across your organization
Harness	Proxy, route, and budget-control all LLM access (Claude Code, Cursor, Codex)
Observatory	Source of truth for all AI expenditure -- costs, tokens, audit trails, metrics
Training	Fine-tune domain LLMs with Unsloth, serve locally at $0/token

Quick Start

pip install sagewai

import asyncio
from sagewai import UniversalAgent

agent = UniversalAgent(name="hello", model="gpt-4o-mini")
print(asyncio.run(agent.chat("What is Sagewai?")))

Three lines to create an agent. Works with GPT-4o, Claude, Gemini, Mistral, Ollama, and 100+ models via LiteLLM.

Install Extras

Extra	What it adds
sagewai[memory]	Milvus, NebulaGraph, Docling, tiktoken
sagewai[intelligence]	Embeddings, entity extraction, language detection
sagewai[postgres]	asyncpg, SQLAlchemy async, Alembic
sagewai[fastapi]	FastAPI + SSE support
sagewai[prometheus]	Prometheus metrics exporter
sagewai[storage]	S3 (boto3) and GCS archival backends
sagewai[all]	Everything above

Examples

21 progressive examples organized by pillar:

SDK (01-08): Hello agent, tools, multi-model, memory, workflows, guardrails, MCP, directives

Harness (09-13): Claude Code proxy, Cursor proxy, Codex proxy, budget enforcement, model routing

Registry (14-16): Register agents, discover agents, agent governance

Training (17-19): Unsloth fine-tune, local LLM routing, domain models

Enterprise (20-21): Fleet workers, full stack setup

See sagewai/examples/ for all examples.

CLI

sagewai init my-project              # scaffold a new project
sagewai doctor                       # check environment health
sagewai agent run my_agent.yaml      # run an agent from config
sagewai harness start                # start the LLM proxy

Documentation

• docs.sagewai.ai -- full documentation
• Getting Started -- quickstart guide
• LLM Harness Guide -- govern AI coding tool costs

Contributing

See CONTRIBUTING.md for development setup, code style, and PR process.

License

AGPL-3.0 -- see LICENSE. Commercial licenses available for organizations that need an alternative to AGPL. See COMMERCIAL_LICENSE.md for details.

Built by Ali Arda Diri.