asadullah48

Agent Starter Kit — AI skill for Claude Code

AI community

A minimal, working Claude-powered AI agent, structured to stay maintainable as tools, memory, APIs, and tests grow -- multi-provider LLM (Claude or free local Ollama), real external-API tools, tests.

How to install Agent Starter Kit

This entry records only its repository, not the path inside it, so there is no exact command to give. Open asadullah48/agent-starter-kit and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Agent Starter Kit does

A minimal, working Claude-powered AI agent, structured to stay maintainable as tools, memory, APIs, and tests grow -- multi-provider LLM (Claude or free local Ollama), real external-API tools, tests, and CI.

Alternatives in AI

Omniget — Udemy & Hotmart course downloader, YouTube downloader (yt-dlp GUI, 1,800+ sites) + desktop app for AI agents 13.9k ★
Opencodex — Universal provider proxy for OpenAI Codex & Claude Code — use any LLM (Claude, Gemini, Grok, DeepSeek, Ollama… 12.3k ★
Zen MCP Server — The power of Claude Code + Gemini / OpenAI / Grok / OpenRouter / Ollama / Custom Model working as one 11.3k ★

README

agent-starter-kit

[![CI](https://github.com/asadullah48/agent-starter-kit/actions/workflows/ci.yml/badge.svg)](https://github.com/asadullah48/agent-starter-kit/actions/workflows/ci.yml) [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)

A minimal, **working** AI agent built on the Claude API -- organized so it stays maintainable once you add more tools, memory, APIs, tests, or teammates, instead of living in one growing Python file. Runs against Claude out of the box, or completely free and local via Ollama.

Live demo

**https://agent-starter-kit-six.vercel.app**

`/health` -- liveness check
`/docs` -- interactive OpenAPI docs (try /chat from here)
POST /chat runs in a safe "demo mode" (explains itself instead of erroring) because no ANTHROPIC_API_KEY is set on the deployment -- a public endpoint with a real key attached would let any visitor spend it. Clone the repo and add your own key (or LLM_PROVIDER=ollama) to try the full agent. Note the deployment is also stateless per-request, so multi-turn memory only works locally, not against the hosted demo.

What's here

One folder = one responsibility:

agent-starter-kit/
??? README.md               Documentation & setup (this file)
??? LICENSE                 MIT
??? requirements.txt         Runtime dependencies
??? requirements-dev.txt      + pytest, for running the test suite
??? .env.example               API keys & environment variables (copy to .env)
??? .gitignore
??? docker-compose.yml         Runs the API in a container
??? Dockerfile
??? vercel.json                 Serverless function config for the live demo
??? main.py                     CLI entry point (chat REPL or --api server)
??? pytest.ini
??? api/index.py                 Vercel entrypoint (re-exports src/api/app.py)
??? .github/workflows/ci.yml     Runs the test suite on every push/PR