llm-coding-agent 0.1a0

2026-07-02 · Source: Simon Willison's Weblog · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, short

Summary

The "llm-coding-agent 0.1a0" was released on July 2nd, 2026, as a new coding agent built upon the LLM library's agent framework. Developed through a Fable 5 experiment, the agent's specification and initial implementation were guided by prompts to Claude Code for web, following a red/green TDD approach. This "slop-alpha" version is available on PyPI, offering both a command-line interface via `uvx` and a Python API through a `CodingAgent` class, which supports models like "gpt-5.5". The agent includes six core tools: `edit_file`, `execute_command`, `list_files`, `read_file`, `search_files`, and `write_file`, enabling comprehensive file and command manipulation. A demonstration showed the agent successfully creating an ASCII art time CLI app, with GPT-5.5 noting SwiftUI's limitations for such an application.

Key takeaway

For AI Engineers developing automated coding tools, this release highlights a practical blueprint for building LLM-powered agents. You can leverage existing LLM libraries as agent frameworks and guide development through TDD, even allowing the LLM to define its own useful APIs. Consider integrating `llm-coding-agent`'s toolset, including file editing and command execution, into your own projects to accelerate development workflows and enhance automation capabilities.

Key insights

LLM-powered coding agents can be effectively built using an agent framework and TDD, even generating their own APIs.

Principles

• LLM frameworks can serve as agent development platforms.
• TDD can guide LLM agent development.
• LLMs can autonomously define useful APIs.

Method

Build an LLM-based coding agent by leveraging an LLM library as an agent framework, prompting for a spec, and iteratively developing with red/green TDD and manual API testing.

In practice

• Use `uvx` to run `llm-coding-agent`.
• Integrate `CodingAgent` for programmatic tasks.
• Define agent tools for file and command ops.

Topics

• LLM Agents
• Coding Agents
• Test-Driven Development
• Python Development
• Agent Tooling
• API Generation

Code references

• simonw/llm-coding-agent
• simonw/python-lib-template-repository

Best for: Machine Learning Engineer, AI Engineer, NLP Engineer, Software Engineer

Related on AIssential

• Components of a Coding Agent: How LLMs Become Software Engineers — Coding agents transform LLMs into dynamic, multi-step problem solvers through structured system design.
• Want to Build AI Agents? Start by Calling an LLM API Yourself — Direct LLM API calls are the fundamental building blocks for all AI agents.
• Step-by-Step Guide to Building AI Agents Using LLMs — LLM-powered AI agents enable advanced automation by understanding language, reasoning, and interacting with systems autonomously.
• Programming with LLM Agents in 2025 — LLM agents, like Open Hands, significantly amplify developer productivity by automating iterative coding and visualization tasks.
• Gemma 4 with Pi Coding Agent & llama.cpp | Build LLM Resource Calculator with NextJS | 🔴 Live — PyCoding Agent with local LLMs can build simple apps but struggles with complex UI overhauls and skill integration.
• How I code with agents, without being 'technical' — AI agents enable non-technical users to build complex software by abstracting traditional coding into a guided, iterative process.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Simon Willison's Weblog.