NiJingzheNiJingzheCommunityยท2 Agent Skills Included

SimpleLLMFunc

Build typed LLM functions and chat agents in Python

Turns LLM calls into typed Python functions using decorators, docstring prompts, and Pydantic return types. Builds multi-turn chat agents with tool calling, persistent code execution, self-editing memory, and parallel sub-agent forks. Removes manual prompt plumbing, JSON parsing, and provider wiring so developers ship reliable LLM features faster.
npx skills add NiJingzhe/SimpleLLMFunc --all -g -y
Available:

Gives your AI agent a step-by-step development workflow for modifying the SimpleLLMFunc framework, covering test-first changes, debugging tips, and key code locations.

All Skills in This Repository (2)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install SimpleLLMFunc?โ–ผ

Run `npx skills add NiJingzhe/SimpleLLMFunc --all -g -y` in your terminal to install all skills in this repository globally.

What does SimpleLLMFunc do?โ–ผ

It lets you define LLM calls as ordinary typed Python functions, where the docstring acts as the prompt and the return type controls structured output parsing.

How do I build a chat agent with tools?โ–ผ

Decorate an async function with @llm_chat, pass a toolkit of @tool functions, and stream results with async for; built-in PyRepl and FileToolset give the agent code execution and file access.

Which model providers does SimpleLLMFunc support?โ–ผ

It works with any OpenAI-compatible chat endpoint or the OpenAI Responses API, configured through a provider.json file with key rotation and rate limiting.

Can agents edit their own memory in SimpleLLMFunc?โ–ผ

Yes. The SelfRef system lets agents remember, forget, compact, and fork their own context at runtime through safe, typed context mutations.

Related Repositories in Software Engineering

View All in Software Engineeringโ†’