Metadata-Version: 2.4
Name: llmsurgery
Version: 0.0.13
Summary: LLM Surgery
Author-email: Jeremy Howard <info@fast.ai>
License: Apache-2.0
Project-URL: Repository, https://github.com/AnswerDotAI/llmsurgery
Project-URL: Documentation, https://AnswerDotAI.github.io/llmsurgery/
Keywords: nbdev
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3 :: Only
Requires-Python: >=3.11
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: fastcore>=2.1.13
Requires-Dist: python-fastllm>=0.0.38
Requires-Dist: aidialog>=0.0.11
Provides-Extra: dev
Requires-Dist: exhash; extra == "dev"
Requires-Dist: openai-codex; extra == "dev"
Requires-Dist: claude-agent-sdk; extra == "dev"
Requires-Dist: json5; extra == "dev"
Requires-Dist: tiktoken; extra == "dev"
Dynamic: license-file

# llmsurgery


<!-- WARNING: THIS FILE WAS AUTOGENERATED! DO NOT EDIT! -->

## Install

``` sh
pip install llmsurgery
```

## What’s here

Dialogs are LLM conversations kept as Jupyter notebooks: notes, runnable code with outputs, and prompt/reply pairs in one editable, diffable document. The dialog data model and toolkit live in [aidialog](https://github.com/AnswerDotAI/aidialog). This library contains the provider-session layers:

- `llmsurgery.compact`: bounded conversation rendering and compaction documents.
- `llmsurgery.ant`: Claude Code session transcripts, including reading, writing, search, curation, compaction, and dialog conversion.
- `llmsurgery.oai`: Codex rollouts, including reading, search, compaction, and dialog conversion.
- `llmsurgery.sess`: one entry point across both hosts, finding a session by id prefix and reading it as a dialog, plus the [`sess2nb`](https://AnswerDotAI.github.io/llmsurgery/sess.html#sess2nb) command.

`llmsurgery.skill` is the single pyskills entry point. It routes agents to `doc(llmsurgery.ant)` for Claude sessions, `doc(llmsurgery.oai)` for Codex sessions, and the compaction module when bounded rendering is needed.
