AGL

AGL runs AI agent workflows against a code repository. A workflow is ordinary Python that asks for a worktree, hands roles to agents and lands what comes back — with fingerprinted replay, git worktree isolation, preflight checks and exit codes supplied by the framework rather than written into the workflow.

This release ships no workflows. The ones AGL runs are the ones you write. agl new <name> writes a workflow into your own workspace under AGL_HOME — a directory holding a module and a pyproject.toml, which runs as it stands — and agl run <name> runs it; agl workflows lists what that workspace declares. AGL reads workflows from there and from nowhere else: the agl.workflows entry-point group is the table key each workflow directory's own pyproject.toml writes, rather than a group a distribution installed beside AGL registers into.

A workflow that imports nothing beyond agl needs no environment of its own and runs as it stands. One declaring third-party dependencies in its own pyproject.toml needs agl sync, which installs what your workspace's workflows declare — their dependencies and never the workflows themselves, so importing a workflow still resolves from the source you are editing. It requires uv on your PATH, and nothing runs it for you: agl new stays offline.

Install

uv tool install agents-gl

The distribution on PyPI is agents-gl; the command it installs and the package you import are both agl.

Python 3.14 or newer

This is the one thing most likely to trip you up. AGL declares requires-python = ">=3.14", so a plain pip install agents-gl on an older interpreter refuses before it downloads anything, and it reports that as not finding a version rather than as your interpreter being too old. uv tool install sidesteps it by fetching a matching interpreter itself.

Status

Early. 0.0.1 is a name claim rather than a release, and the surface may still change.

Licence

MIT — see LICENSE.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

agents_gl-0.0.2.tar.gz (144.8 kB view details)

Uploaded Sep 6, 2026 Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files.

agents_gl-0.0.2-py3-none-any.whl (169.5 kB view details)

Uploaded Sep 6, 2026 Python 3

File details

Details for the file agents_gl-0.0.2.tar.gz.

File metadata

Download URL: agents_gl-0.0.2.tar.gz
Upload date: Sep 6, 2026
Size: 144.8 kB
Tags: Source
Uploaded using Trusted Publishing? Yes
Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for agents_gl-0.0.2.tar.gz
Algorithm	Hash digest
SHA256	`ac178499201c14bd659f85001547a64e957c6d870022301f31989f6c36ac2812`
MD5	`2f35d44aa771c74c310383a62d19c76d`
BLAKE2b-256	`ba95912a6206c3532beb1c899e3289c935b5b83605a47d0b68ff3d1cbc1b3e19`

See more details on using hashes here.

Provenance

The following attestation bundles were made for agents_gl-0.0.2.tar.gz:

Publisher: publish.yml on jashioq/AGL

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Statement:
- Statement type: https://in-toto.io/Statement/v1
- Predicate type: https://docs.pypi.org/attestations/publish/v1
- Subject name: agents_gl-0.0.2.tar.gz
- Subject digest: ac178499201c14bd659f85001547a64e957c6d870022301f31989f6c36ac2812
- Sigstore transparency entry: 2733654598
- Sigstore integration time: Sep 6, 2026
Source repository:
- Permalink: jashioq/AGL@16ba2f6ba56faf853a0af6db767332f05a5eaebb
- Branch / Tag: refs/heads/main
- Owner: https://github.com/jashioq
- Access: public
Publication detail:
- Token Issuer: https://token.actions.githubusercontent.com
- Runner Environment: github-hosted
- Publication workflow: publish.yml@16ba2f6ba56faf853a0af6db767332f05a5eaebb
- Trigger Event: workflow_dispatch