messygit 0.2.0__tar.gz → 0.3.0__tar.gz

messygit-0.3.0/.github/workflows/publish.yml

@@ -0,0 +1,27 @@
+name: Publish to PyPI
+
+on:
+  release:
+    types: [published]
+
+jobs:
+  publish:
+    runs-on: ubuntu-latest
+    permissions:
+      id-token: write
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.13"
+
+      - name: Install build tools
+        run: pip install build
+
+      - name: Build package
+        run: python -m build
+
+      - name: Publish to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1

messygit-0.3.0/.vscode/settings.json

@@ -0,0 +1 @@
+{}

messygit-0.3.0/ARCHITECTURE.md

@@ -0,0 +1,64 @@
+# Architecture
+
+This document describes the purpose of each file in the `messygit` project.
+
+## Root
+
+| File                              | Purpose                                                                                                                        |
+| --------------------------------- | ------------------------------------------------------------------------------------------------------------------------------ |
+| `pyproject.toml`                  | Package metadata, dependencies (`anthropic`, `click`), build system (hatchling), and the `messygit` console script entrypoint. |
+| `README.md`                       | User-facing documentation: install, usage, commands, and development instructions.                                             |
+| `.gitignore`                      | Keeps `.venv/`, `__pycache__/`, `dist/`, and `*.egg-info/` out of version control.                                             |
+| `.github/workflows/publish.yml`   | GitHub Actions workflow for publishing the package.                                                                             |
+
+## `messygit/` (Python package)
+
+| File          | Purpose                                                                                                                                                                                                                                                                 |
+| ------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| `__init__.py` | Marks the directory as a Python package (empty).                                                                                                                                                                                                                        |
+| `cli.py`      | Interactive REPL entrypoint. Displays an ASCII banner, runs a prompt loop, and dispatches commands: `add`, `commit`, `push`, `config`, `show`, `suggest`, `help`, `quit`/`exit`. Orchestrates all other modules. Includes a threaded `Spinner` for loading animations.   |
+| `git.py`      | All subprocess calls to `git`. Reads staged diffs (`git diff --cached -U0`), parses them into a compact changed-lines format, filters noise files, handles the large-diff fallback (stat summary + top-N most-changed files), and runs `git add`, `git commit`, and `git push`. |
+| `llm.py`      | Anthropic SDK integration. Creates the client with the resolved API key, calls `messages.create`, extracts the text response, and maps SDK exceptions (`AuthenticationError`, `PermissionDeniedError`, `BadRequestError`, billing 402) into user-friendly error classes. Also exposes helpers reused by `agent/agent.py`. |
+| `config.py`   | API key storage and resolution. Reads/writes `~/.messygit/config.json`, checks the `ANTHROPIC_API_KEY` env var, validates keys are non-empty, masks keys for display, and defines all user-facing error messages and exception classes.                                 |
+| `prompts.py`  | System prompts and user prompt builder. Contains the Conventional Commits instructions (`COMMIT_SYSTEM_PROMPT`), the suggestion agent instructions (`SUGGESTION_SYSTEM_PROMPT`), and `build_user_prompt` which wraps staged changes into the user message sent to Claude. |
+
+## `messygit/agent/` (agentic tool-use subpackage)
+
+| File       | Purpose                                                                                                                                                                                                                           |
+| ---------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| `tool.py`  | `Tool` dataclass wrapping a Python callable with a name, description, and parameter schema. Provides `run()` to invoke the function and `to_schema()` to emit an Anthropic-compatible tool definition.                            |
+| `tools.py` | Concrete tool instances available to agents: `run_git_tool` (allowlisted read-only git commands), `read_file_tool`, `list_directory_tool`, and `search_code_tool` (`git grep`).                                                   |
+| `agent.py` | Generic agentic loop. Sends messages to Claude with tool definitions, processes `tool_use` response blocks by dispatching to matching `Tool` instances, appends results, and iterates up to `max_iterations`. Handles API errors identically to `llm.py`. |
+
+## Data flow
+
+```
+User runs `messygit` → interactive REPL
+       │
+       ├── "add <files>"  ──► git.py (git add)
+       ├── "push"         ──► git.py (git push)
+       ├── "config <key>" ──► config.py (save API key)
+       ├── "show"         ──► config.py (display masked key)
+       │
+       ├── "commit" ──► git.py (read staged diff, apply token budget)
+       │       │
+       │       ▼
+       │   llm.py (send context to Claude)
+       │       │       │
+       │       │       ├── config.py (resolve API key)
+       │       │       └── prompts.py (COMMIT_SYSTEM_PROMPT + user prompt)
+       │       ▼
+       │   cli.py (display message, prompt Y/n/e)
+       │       │
+       │       ▼
+       │   git.py (git commit -m "...")
+       │
+       └── "suggest" ──► agent/agent.py (agentic loop)
+               │               │
+               │               ├── config.py (resolve API key)
+               │               ├── prompts.py (SUGGESTION_SYSTEM_PROMPT)
+               │               └── agent/tools.py (run_git, read_file, list_directory)
+               ▼
+           cli.py (display suggestion)
+```
+

messygit-0.3.0/PKG-INFO

@@ -0,0 +1,214 @@
+Metadata-Version: 2.4
+Name: messygit
+Version: 0.3.0
+Summary: An AI-powered interactive git CLI with agentic tools for commits, code suggestions, and workflow automation.
+License-Expression: MIT
+Requires-Python: >=3.10
+Requires-Dist: anthropic>=0.39.0
+Requires-Dist: click>=8.0
+Requires-Dist: rich>=13.0
+Description-Content-Type: text/markdown
+
+# messygit
+
+**messygit** is an interactive CLI that turns messy git workflows into clean Conventional Commits — stage, commit, push, and get AI-powered project suggestions, all from one interface powered by [Claude](https://www.anthropic.com/api).
+
+## Why use it
+
+- **Interactive REPL** — one command drops you into a persistent session where you can stage, commit, push, and more without leaving.
+- **AI commit messages** — sends your staged diff to Claude and suggests a clean Conventional Commits subject line.
+- **Project suggestions** — an AI agent inspects your repo and recommends concrete next steps.
+- **Token usage & cost** — tracks the tokens each session uses and shows a rough cost estimate, with a one-command jump to billing.
+- **Themed UI** — a colored startup animation and prompt you can recolor with the `theme` command.
+- **Safe by default** — only the staged diff is sent to the model. Your API key is never printed in full.
+
+## Requirements
+
+- **Python** 3.10 or newer
+- **Git** (run inside a repository)
+- An **Anthropic API key** with access to the Messages API
+
+## Installation
+
+```bash
+pip install messygit
+```
+
+### Install from source
+
+```bash
+cd messygit
+python -m venv .venv
+source .venv/bin/activate   # Windows: .venv\Scripts\activate
+pip install -e .
+```
+
+## API key
+
+messygit resolves the key in this order:
+
+1. Environment variable **`ANTHROPIC_API_KEY`**
+2. Config file **`~/.messygit/config.json`**
+
+You can set the key from within the messygit interface:
+
+```
+messygit > config YOUR_ANTHROPIC_API_KEY
+```
+
+## Usage
+
+```bash
+messygit
+```
+
+This drops you into the interactive interface — an animated banner followed by a
+status dashboard, then the prompt:
+
+```
+ mmm    mmmm  eeeeeee  sssssss  sssssss  yy   yy  ggggggg  ii  tttttttt
+ mm mm mm mm  ee       ss       ss        yy yy   gg       ii     tt
+ mm  mmm  mm  eeeee    sssssss  sssssss    yy     gg  ggg  ii     tt
+ mm       mm  ee            ss       ss    yy     gg   gg  ii     tt
+ mm       mm  eeeeeee  sssssss  sssssss    yy     ggggg    ii     tt
+
+  repo     messygit  ⎇ main
+  status   3 staged · 2 modified
+  api key  sk-ant-a...x3f2 (config)
+  model    Haiku 4.5  $1 in · $5 out / 1M
+  tokens   0 used this session
+  ────────────────────────────────────────────────────────────
+  Type help for commands · quit to exit
+
+messygit (main) ❯
+```
+
+### Commands
+
+Commands are grouped on the `help` screen:
+
+**git**
+
+| Command | Description |
+|---------|-------------|
+| `add <file>` or `add .` | Stage files for commit |
+| `commit` | Generate an AI commit message from staged changes, then commit / cancel / edit |
+| `push` | Push commits to remote |
+
+**messyagent**
+
+| Command | Description |
+|---------|-------------|
+| `suggest` | Get AI-powered next-step suggestions for your project |
+
+**account**
+
+| Command | Description |
+|---------|-------------|
+| `config <key>` | Save your Anthropic API key to `~/.messygit/config.json` |
+| `show` | Display a masked API key and its source |
+| `model` or `model <name>` | Switch the Claude model (run `model` to list models and pricing) |
+| `tokens` | Show this session's token usage and estimated cost, and open the billing console |
+
+**app**
+
+| Command | Description |
+|---------|-------------|
+| `theme` or `theme <name>` | Change the UI color (run `theme` to list presets) |
+| `help` | List available commands |
+| `quit` / `exit` | Exit messygit |
+
+### Typical flow
+
+```
+messygit > add .
+Staged everything
+
+messygit > commit
+feat(cli): add interactive REPL with ASCII banner
+Commit with this message? [y/n/e] y
+
+messygit > push
+```
+
+> Tip: run `suggest` for AI next-step ideas, or `theme` to recolor the UI.
+
+### Commit message style
+
+The model follows **Conventional Commits**: `type(scope): description`
+
+Allowed types: `feat`, `fix`, `docs`, `style`, `refactor`, `test`, `chore`. Subjects are one line, imperative, lowercase, no trailing period.
+
+### Token usage & cost
+
+messygit tracks the tokens used by AI commands (`commit`, `suggest`) for the current session and shows a running total after each call. Run `tokens` for a breakdown and a one-key jump to the Anthropic billing console:
+
+```
+messygit > tokens
+╭─ token usage ─────────────────────────────╮
+│ used     8,420 tokens                     │
+│          6,200 in · 2,220 out             │
+│ requests 2                                │
+│ est cost ≈ $0.02                          │
+╰───────────────────────────────────────────╯
+```
+
+The cost is a **rough estimate** at the selected model's [pricing](https://www.anthropic.com/pricing). The Anthropic API does not expose remaining account credits, so usage and cost are measured per session only.
+
+### Choosing a model
+
+messygit defaults to **Claude Haiku 4.5** (fast and cheap). Use `model` to list the available models and their token pricing, or `model <name>` to switch (your choice persists in `~/.messygit/config.json`):
+
+```
+messygit > model
+Available models — usage: model <name>
+  ● haiku  Haiku 4.5   $1 in · $5 out / 1M  (current)
+    sonnet Sonnet 4.6  $3 in · $15 out / 1M
+    opus   Opus 4.8    $5 in · $25 out / 1M
+
+messygit > model opus
+! Opus 4.8 costs more than Haiku 4.5 ($5 in · $25 out / 1M vs $1 in · $5 out / 1M). Token usage will be billed at the higher rate.
+Switch anyway? [y/N]:
+```
+
+Switching to a **more expensive** model prompts for confirmation first. The session cost estimate prices each request at the model used for it, so it stays accurate even if you switch mid-session.
+
+## Publishing to PyPI
+
+This project uses GitHub Actions with [PyPI trusted publishing](https://docs.pypi.org/trusted-publishers/) — no API tokens needed in your repo.
+
+### One-time setup
+
+1. Go to your project on [pypi.org](https://pypi.org/manage/project/messygit/settings/publishing/)
+2. Add a **Trusted Publisher**:
+   - **Owner:** your GitHub username
+   - **Repository:** `messygit`
+   - **Workflow name:** `publish.yml`
+   - **Environment:** leave blank
+
+### To release a new version
+
+1. Bump `version` in `pyproject.toml`
+2. Commit and push
+3. Create a GitHub release:
+
+```bash
+git tag v0.2.0
+git push origin v0.2.0
+```
+
+4. Go to GitHub → Releases → Draft a new release → select the tag → Publish
+
+The workflow at `.github/workflows/publish.yml` will automatically build and upload to PyPI.
+
+### Manual publish (without CI)
+
+```bash
+rm -rf dist/
+python -m build
+twine upload dist/*
+```
+
+## License
+
+MIT (see `pyproject.toml`).

messygit-0.3.0/README.md

@@ -0,0 +1,203 @@
+# messygit
+
+**messygit** is an interactive CLI that turns messy git workflows into clean Conventional Commits — stage, commit, push, and get AI-powered project suggestions, all from one interface powered by [Claude](https://www.anthropic.com/api).
+
+## Why use it
+
+- **Interactive REPL** — one command drops you into a persistent session where you can stage, commit, push, and more without leaving.
+- **AI commit messages** — sends your staged diff to Claude and suggests a clean Conventional Commits subject line.
+- **Project suggestions** — an AI agent inspects your repo and recommends concrete next steps.
+- **Token usage & cost** — tracks the tokens each session uses and shows a rough cost estimate, with a one-command jump to billing.
+- **Themed UI** — a colored startup animation and prompt you can recolor with the `theme` command.
+- **Safe by default** — only the staged diff is sent to the model. Your API key is never printed in full.
+
+## Requirements
+
+- **Python** 3.10 or newer
+- **Git** (run inside a repository)
+- An **Anthropic API key** with access to the Messages API
+
+## Installation
+
+```bash
+pip install messygit
+```
+
+### Install from source
+
+```bash
+cd messygit
+python -m venv .venv
+source .venv/bin/activate   # Windows: .venv\Scripts\activate
+pip install -e .
+```
+
+## API key
+
+messygit resolves the key in this order:
+
+1. Environment variable **`ANTHROPIC_API_KEY`**
+2. Config file **`~/.messygit/config.json`**
+
+You can set the key from within the messygit interface:
+
+```
+messygit > config YOUR_ANTHROPIC_API_KEY
+```
+
+## Usage
+
+```bash
+messygit
+```
+
+This drops you into the interactive interface — an animated banner followed by a
+status dashboard, then the prompt:
+
+```
+ mmm    mmmm  eeeeeee  sssssss  sssssss  yy   yy  ggggggg  ii  tttttttt
+ mm mm mm mm  ee       ss       ss        yy yy   gg       ii     tt
+ mm  mmm  mm  eeeee    sssssss  sssssss    yy     gg  ggg  ii     tt
+ mm       mm  ee            ss       ss    yy     gg   gg  ii     tt
+ mm       mm  eeeeeee  sssssss  sssssss    yy     ggggg    ii     tt
+
+  repo     messygit  ⎇ main
+  status   3 staged · 2 modified
+  api key  sk-ant-a...x3f2 (config)
+  model    Haiku 4.5  $1 in · $5 out / 1M
+  tokens   0 used this session
+  ────────────────────────────────────────────────────────────
+  Type help for commands · quit to exit
+
+messygit (main) ❯
+```
+
+### Commands
+
+Commands are grouped on the `help` screen:
+
+**git**
+
+| Command | Description |
+|---------|-------------|
+| `add <file>` or `add .` | Stage files for commit |
+| `commit` | Generate an AI commit message from staged changes, then commit / cancel / edit |
+| `push` | Push commits to remote |
+
+**messyagent**
+
+| Command | Description |
+|---------|-------------|
+| `suggest` | Get AI-powered next-step suggestions for your project |
+
+**account**
+
+| Command | Description |
+|---------|-------------|
+| `config <key>` | Save your Anthropic API key to `~/.messygit/config.json` |
+| `show` | Display a masked API key and its source |
+| `model` or `model <name>` | Switch the Claude model (run `model` to list models and pricing) |
+| `tokens` | Show this session's token usage and estimated cost, and open the billing console |
+
+**app**
+
+| Command | Description |
+|---------|-------------|
+| `theme` or `theme <name>` | Change the UI color (run `theme` to list presets) |
+| `help` | List available commands |
+| `quit` / `exit` | Exit messygit |
+
+### Typical flow
+
+```
+messygit > add .
+Staged everything
+
+messygit > commit
+feat(cli): add interactive REPL with ASCII banner
+Commit with this message? [y/n/e] y
+
+messygit > push
+```
+
+> Tip: run `suggest` for AI next-step ideas, or `theme` to recolor the UI.
+
+### Commit message style
+
+The model follows **Conventional Commits**: `type(scope): description`
+
+Allowed types: `feat`, `fix`, `docs`, `style`, `refactor`, `test`, `chore`. Subjects are one line, imperative, lowercase, no trailing period.
+
+### Token usage & cost
+
+messygit tracks the tokens used by AI commands (`commit`, `suggest`) for the current session and shows a running total after each call. Run `tokens` for a breakdown and a one-key jump to the Anthropic billing console:
+
+```
+messygit > tokens
+╭─ token usage ─────────────────────────────╮
+│ used     8,420 tokens                     │
+│          6,200 in · 2,220 out             │
+│ requests 2                                │
+│ est cost ≈ $0.02                          │
+╰───────────────────────────────────────────╯
+```
+
+The cost is a **rough estimate** at the selected model's [pricing](https://www.anthropic.com/pricing). The Anthropic API does not expose remaining account credits, so usage and cost are measured per session only.
+
+### Choosing a model
+
+messygit defaults to **Claude Haiku 4.5** (fast and cheap). Use `model` to list the available models and their token pricing, or `model <name>` to switch (your choice persists in `~/.messygit/config.json`):
+
+```
+messygit > model
+Available models — usage: model <name>
+  ● haiku  Haiku 4.5   $1 in · $5 out / 1M  (current)
+    sonnet Sonnet 4.6  $3 in · $15 out / 1M
+    opus   Opus 4.8    $5 in · $25 out / 1M
+
+messygit > model opus
+! Opus 4.8 costs more than Haiku 4.5 ($5 in · $25 out / 1M vs $1 in · $5 out / 1M). Token usage will be billed at the higher rate.
+Switch anyway? [y/N]:
+```
+
+Switching to a **more expensive** model prompts for confirmation first. The session cost estimate prices each request at the model used for it, so it stays accurate even if you switch mid-session.
+
+## Publishing to PyPI
+
+This project uses GitHub Actions with [PyPI trusted publishing](https://docs.pypi.org/trusted-publishers/) — no API tokens needed in your repo.
+
+### One-time setup
+
+1. Go to your project on [pypi.org](https://pypi.org/manage/project/messygit/settings/publishing/)
+2. Add a **Trusted Publisher**:
+   - **Owner:** your GitHub username
+   - **Repository:** `messygit`
+   - **Workflow name:** `publish.yml`
+   - **Environment:** leave blank
+
+### To release a new version
+
+1. Bump `version` in `pyproject.toml`
+2. Commit and push
+3. Create a GitHub release:
+
+```bash
+git tag v0.2.0
+git push origin v0.2.0
+```
+
+4. Go to GitHub → Releases → Draft a new release → select the tag → Publish
+
+The workflow at `.github/workflows/publish.yml` will automatically build and upload to PyPI.
+
+### Manual publish (without CI)
+
+```bash
+rm -rf dist/
+python -m build
+twine upload dist/*
+```
+
+## License
+
+MIT (see `pyproject.toml`).

{messygit-0.2.0 → messygit-0.3.0}/messygit/agent/agent.py

@@ -15,8 +15,9 @@ from ..config import (
     resolve_api_key,
 )
 from ..llm import _is_insufficient_balance_or_billing, _insufficient_balance_user_message, _text_from_message
+from ..models import current_model
+from ..usage import SESSION_USAGE
 
-DEFAULT_MODEL = "claude-haiku-4-5-20251001"
 DEFAULT_MAX_TOKENS = 4096
 
 class Agent:
@@ -29,19 +30,21 @@ class Agent:
     def run(self, user_input: str) -> str:
         """Run the agent."""
         client = Anthropic(api_key=resolve_api_key())
+        model = current_model()
         messages = []
         try:
             messages.append({"role": "user", "content": user_input})
             response = None
             for i in range(self.max_iterations):
                 response = client.messages.create(
-                    model=DEFAULT_MODEL,
+                    model=model.id,
                     max_tokens=DEFAULT_MAX_TOKENS,
                     tools=[t.to_schema() for t in self.tools],
                     tool_choice={"type": "auto"},
                     system=self.system_prompt,
                     messages=messages,
                 )
+                SESSION_USAGE.record(response.usage, model)
                 messages.append({"role": "assistant", "content": response.content})
 
                 tool_use_blocks = [b for b in response.content if b.type == "tool_use"]