Agent Worktree Manager
Working on several branches or with coding agents means keeping track of both
source checkouts and their Python environments. awm pairs Git worktrees with
isolated environments and provides a terminal UI to browse and clean them up
across projects.
Start with an environment that already works. AWM reproduces its installed dependencies and points selected editable packages at new worktrees. Other editable packages keep their original source links. Install AWM once, separately from your project environments.
Supports Python 3.11–3.14 on macOS and Linux; Git is required. Standard venv/pip is the default; uv and conda backends are optional. AWM and your project environments can use different Python versions.
Installation
From PyPI, once the first release is published, choose either:
uv tool install agent-worktree-manager
# Or install into your chosen Python environment:
python -m pip install agent-worktree-manager
To install from a source checkout now, run at this repository's root:
uv tool install --editable .
# Or:
python -m pip install -e .
To activate a sandbox in your current terminal from the TUI, enable shell integration:
eval "$(awm shell-init zsh)" # use bash for Bash
Add the same line to ~/.zshrc (or ~/.bashrc) to enable it in new terminals.
Then press s on a sandbox to close AWM and activate its environment.
Register your first project
From your project's Git checkout, register its existing Python environment:
awm init --name my-project --base /absolute/path/to/base-env --repo app=.
awm doctor
awm create my-task --repo app
awm run my-task -- python --version
awm
--base is the environment directory, not its Python executable. Use
--backend uv for a uv base or --backend conda for a conda base. Repeat
--repo ALIAS=PATH to include more repositories. Registration writes awm.toml
and machine-specific settings in git-ignored .awm/, and adds the project to
your global overview. From elsewhere, select it with awm --project my-project.
Use with coding agents
The bundled awm skill guides agents through creating, resuming, inspecting,
and cleaning up worktrees with their environments. Install it for either agent:
# Codex
awm skill --install "$HOME/.agents/skills/awm"
# Claude Code
awm skill --install "$HOME/.claude/skills/awm"
In Codex, select it through /skills or mention $awm. In Claude Code, use
/awm. For example: “create a sandbox for fixing the login tests and work there.”
See the agent guide for custom locations and updating a skill.
Contributing
Please follow Conventional Commits
for commit messages, for example feat(tui): add sandbox shell shortcut or
fix(env): preserve editable paths. These messages drive automated versioning
and release notes.
License
BSD 3-Clause — Copyright (c) 2026, Jan Wyzula.
Documentation
Metadata
Release files for agent-worktree-manager 0.2.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| agent_worktree_manager-0.2.0.tar.gz | 64.4 kB | With provenance | Details for agent_worktree_manager-0.2.0.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| agent_worktree_manager-0.2.0-py3-none-any.whl | Python 3 | none | any | Details for agent_worktree_manager-0.2.0-py3-none-any.whl |
Total release size: 112.4 kB
Release files / agent_worktree_manager-0.2.0.tar.gz
| Download URL | agent_worktree_manager-0.2.0.tar.gz |
|---|---|
| Size | 64.4 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
bdf253e0a1e108005dcf1a9bcb9ade55421ffd0d39fdd092f633acf365a80407
|
|
BLAKE2b-256 checksum How to use checksums |
6fe585e31ea0882f81353f981239c412a75a26b3621e6808147a6e2c955b1012
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Sep 10, 2026.
Transparency logRelease files / agent_worktree_manager-0.2.0-py3-none-any.whl
| Download URL | agent_worktree_manager-0.2.0-py3-none-any.whl |
|---|---|
| Size | 48.0 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
12eae8b8213b1dfb6577f6e65a8b538478b4e161de8f1d2cc60d87385480bf58
|
|
BLAKE2b-256 checksum How to use checksums |
6dd5e185d649aac1efe9457cef7cdb5c997d840eae6dce96d79632e3e62d845d
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Sep 10, 2026.
Transparency logSecurity for agent-worktree-manager 0.2.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
Full provenance
All 2 files in this release have been uploaded with provenance metadata.
Attestation for agent_worktree_manager-0.2.0.tar.gz
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Sep 10, 2026.
Transparency logA note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.