llm-coding-agent 0.1a0: CLI & Python API Coding Agent

llm-coding-agent 0.1a0 by Simon Willison is a simple coding agent built on his LLM library (evolved into an Agent framework). It provides CLI commands (e.g., `uvx --prerelease=allow --with llm-coding-agent llm code`) and a Python API (e.g., `CodingAgent(model="gpt-5.5", root="/path", approve=True).run("Fix the failing test...")`), supporting yolo mode (`--yolo`) and allowlist mode (`--allow "pytest*"`). Built-in tools include: `edit_file` (precise string replacement with diff output), `execute_command` (runs commands with 600s timeout), `list_files` (glob listing, skipping hidden dirs and .gitignore), `read_file` (chunked reading with offset/limit), `search_files` (regex search), and `write_file` (creates or overwrites files, auto-creating parent dirs). Developers can quickly build coding assistants like Claude Code, reducing repetitive work, and integrate via Python API into existing automation pipelines.