---
title: "Coding agent built on LLM"
canonical: "https://agenticfeats.com/c/coding-agent-built-on-llm-claude-code/"
agents: ["Claude Code"]
evidence_grade: "A"
outcome: "completed"
job_type: "Coding"
roles: ["Developer"]
source_url: "https://simonwillison.net/2026/Jul/2/llm-coding-agent"
evidence_url: "https://github.com/simonw/llm-coding-agent/commits/0.1a0"
published: "2026-10-13"
updated: "2026-10-13"
---

> From Agentic Feats: sourced examples of AI agents at work. Map of the site for agents: https://agenticfeats.com/llms.txt

# Coding agent built on LLM

Claude Code on the web wrote a spec, then built and tested a coding agent library on LLM using red/green TDD.

Result: Shipped llm-coding-agent 0.1a0 to PyPI with six file and shell tools, a CLI and a Python API; it built a working ASCII-art clock app.

## How it was set up

Claude Code for web on a Python library template repo, depending on the latest llm alpha. Manual testing used an OpenAI API key with GPT-5.5.

## Worth knowing

Released as an early alpha the author calls slop-alpha; the first attempt is described as pretty good, not finished.

## Prompt

```text
Write a spec.md for this project. It will depend on [library] and implement [kind of tool] with [list of capabilities]. Then commit the spec and build it using red/green TDD in sensible commits, each with passing tests and updated docs. Manually test occasionally using [API key] in your environment.
```

## Evidence note

Public repo with spec, README, commit history, and demo transcript gist

## Sources

- Original: https://simonwillison.net/2026/Jul/2/llm-coding-agent
- Evidence: https://github.com/simonw/llm-coding-agent/commits/0.1a0

## The receipt

- Agent: Claude Code
- Evidence: Shown (grade A, artifact linked)
- Outcome: Completed
- Job type: Coding
- Roles: Developer
- Source: Blog
- Added: 2026-10-13