emmanuelerogian723-alt/stew-python

Africa's #1 AI Agent API Python SDK — 60+ skills, web research, document generation, code execution, OCR, 100-agent swarm. Zero dependencies. Python 3.8+. # stew-ai 🥘 (Python) **Africa's #1 AI Agent API** — 60+ skills, web research, document generation, code execution, OCR, and a 100-agent swarm. Zero dependencies. Python 3.8+. > **S.T.E.W** = **S**mart **T**ask **E**xecution **W**orker One SDK. Every skill your app needs. Chat with 6 AI providers (automatic failover), search the web, generate PDF/DOCX/XLSX/PPTX, run Python code, OCR images, and spawn 100 AI agents — all from a single Python package. Zero dependencies (uses stdlib `urllib` only). 🌐 **Live API**: stew-agent.onrender.com · 📚 **Docs**: stew-agent.onrender.com --- ## Install ```bash pip install stew-ai ``` This installs both the Python SDK and the `stew` CLI command. --- ## Quick Start (3 lines) ```python from stew_ai import Stew stew = Stew(api_key="stew_your_api_key_here") response = stew.chat.send("Hello from Africa!") print(response["response"]) ``` Get a free API key at stew-agent.onrender.com — no credit card needed. --- ## CLI Usage After `pip install stew-ai`, you get the `stew` command: ```bash # Save your API key stew login stew_your_api_key_here # Chat stew chat "What is the capital of Nigeria?" stew chat "Latest news in Lagos" --web stew chat "Write a poem" --json # Search stew search "top Nigerian fintechs 2026" # Skills (59 available) stew skills stew skills run generate_cv '{"name":"Emmanuel","role":"Developer"}' # Documents stew doc pdf '{"title":"Report","content":"Hello"}' --output report.pdf # Fine-tune stew finetune --persona doctor --instructions "Always cite NHS guidelines" # Status stew status # Piping cat file.txt | stew chat "Summarize this" ``` --- ## SDK Methods ### Chat ```python stew = Stew(api_key="stew_xxx") # Basic chat response = stew.chat.send("What is 2+2?") print(response["response"]) # Chat with web search response = stew.chat.send("Latest news in Nigeria?", web_search=True) # Chat with conversation context response = stew.chat.send("Follow up question", …