[](https://pypi.org/project/agilab)
[](https://pypi.org/project/agilab/)
[](https://opensource.org/licenses/BSD-3-Clause)
[]()
[](https://thalesgroup.github.io/agilab/tests.svg)
[](https://thalesgroup.github.io/agilab/coverage.svg)
[](https://github.com/ThalesGroup/agilab)
[]()
[](https://thalesgroup.github.io/agilab)
[](https://orcid.org/0009-0003-5375-368X)
# AGILAB Open Source Project
AGILAB [BSD license](https://github.com/ThalesGroup/agilab/blob/main/LICENSE) is a PyCharm‑integrated AI experimentation lab for engineering. It’s designed to help engineers quickly experiment with AI‑driven methods using the provided PyCharm run configurations and workflows. Note: full workflow assumes PyCharm IDE.
See [documentation](https://thalesgroup.github.io/agilab).
See also: CHANGELOG.md for recent changes.
## Install and Execution for end users
Quick run (no setup):
```bash
uvx -p 3.13 agilab
```
Managed workspace (project folder):
```bash
mkdir agi-space && cd agi-space
uv init --bare --no-workspace
uv add agilab
uv run agilab
```
## Install for developers
<details open>
<summary>
<strong> Linux and MacOs </strong>
</summary>
```bash
git clone https://github.com/ThalesGroup/agilab
cd agilab
./install.sh --openai-api-key "sk-your-api-key" --cluster-ssh-credentials "username[:password]"
```
</details>
<details>
<summary>
<strong>Windows</strong>
</summary>
```powershell
git clone https://github.com/ThalesGroup/agilab
cd agilabpush it
powershell.exe -ExecutionPolicy Bypass -File .\install.ps1 --openai-api-key "sk-your-api-key"
```
</details>
## AGILab Execution
### Linux and MacOS and Windows:
```bash
cd agilab/src/agilab
uv run agilab
```
## Notes for developers
- AgiEnv is a singleton. Use instance attributes (`env.apps_dir`, `env.logger`, etc.).
Class attribute reads (e.g., `AgiEnv.apps_dir`) proxy to the singleton when initialised;
methods/properties are not shadowed. A few helpers are pre‑init safe
(`AgiEnv.set_env_var`, `AgiEnv.read_agilab_path`, `AgiEnv._build_env`, `AgiEnv.log_info`).
- Environment flags (replaces legacy `install_type`):
- `env.is_source_env`: true when running from a source checkout.
- `env.is_worker_env`: true in worker-only contexts (e.g., `wenv/*_worker`).
- `env.is_local_worker`: helper flag for home‑scoped worker layouts.
- App constructors (templates + flight_project) ignore unknown kwargs when constructing
their Pydantic `Args` models. This preserves strict validation while making constructors
resilient to incidental extras. Configure verbosity via `AgiEnv(verbose=…)` or logging,
not via app `Args`.
Raw data
{
"_id": null,
"home_page": null,
"name": "agilab",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.11",
"maintainer_email": null,
"keywords": "jupyter, mlflow, asyncio, dask, rapids, streamlit, distributed, cython, cluster, dataframe, dataset, load balancing, gen ai, copilot, agi, pycharm, datascience",
"author": null,
"author_email": "Jean-Pierre Morard <focus@thalesgroup.com>",
"download_url": null,
"platform": null,
"description": "[](https://pypi.org/project/agilab)\n[](https://pypi.org/project/agilab/)\n[](https://opensource.org/licenses/BSD-3-Clause)\n[]()\n[](https://thalesgroup.github.io/agilab/tests.svg)\n[](https://thalesgroup.github.io/agilab/coverage.svg)\n[](https://github.com/ThalesGroup/agilab)\n[]()\n[](https://thalesgroup.github.io/agilab)\n[](https://orcid.org/0009-0003-5375-368X)\n\n\n# AGILAB Open Source Project\n\nAGILAB [BSD license](https://github.com/ThalesGroup/agilab/blob/main/LICENSE) is a PyCharm\u2011integrated AI experimentation lab for engineering. It\u2019s designed to help engineers quickly experiment with AI\u2011driven methods using the provided PyCharm run configurations and workflows. Note: full workflow assumes PyCharm IDE.\nSee [documentation](https://thalesgroup.github.io/agilab).\n\nSee also: CHANGELOG.md for recent changes.\n\n## Install and Execution for end users\n\nQuick run (no setup):\n\n```bash\nuvx -p 3.13 agilab\n```\n\nManaged workspace (project folder):\n\n```bash\nmkdir agi-space && cd agi-space\nuv init --bare --no-workspace\nuv add agilab\nuv run agilab\n```\n\n## Install for developers\n\n<details open> \n<summary>\n <strong> Linux and MacOs </strong>\n</summary>\n\n```bash\ngit clone https://github.com/ThalesGroup/agilab\ncd agilab\n./install.sh --openai-api-key \"sk-your-api-key\" --cluster-ssh-credentials \"username[:password]\"\n```\n</details>\n\n<details> \n<summary>\n <strong>Windows</strong>\n</summary>\n\n```powershell\ngit clone https://github.com/ThalesGroup/agilab\ncd agilabpush it\npowershell.exe -ExecutionPolicy Bypass -File .\\install.ps1 --openai-api-key \"sk-your-api-key\"\n```\n</details>\n\n## AGILab Execution\n\n### Linux and MacOS and Windows:\n\n```bash\ncd agilab/src/agilab\nuv run agilab\n```\n\n## Notes for developers\n\n- AgiEnv is a singleton. Use instance attributes (`env.apps_dir`, `env.logger`, etc.).\n Class attribute reads (e.g., `AgiEnv.apps_dir`) proxy to the singleton when initialised;\n methods/properties are not shadowed. A few helpers are pre\u2011init safe\n (`AgiEnv.set_env_var`, `AgiEnv.read_agilab_path`, `AgiEnv._build_env`, `AgiEnv.log_info`).\n\n- Environment flags (replaces legacy `install_type`):\n - `env.is_source_env`: true when running from a source checkout.\n - `env.is_worker_env`: true in worker-only contexts (e.g., `wenv/*_worker`).\n - `env.is_local_worker`: helper flag for home\u2011scoped worker layouts.\n\n- App constructors (templates + flight_project) ignore unknown kwargs when constructing\n their Pydantic `Args` models. This preserves strict validation while making constructors\n resilient to incidental extras. Configure verbosity via `AgiEnv(verbose=\u2026)` or logging,\n not via app `Args`.\n",
"bugtrack_url": null,
"license": null,
"summary": "AGILAB is a PyCharm\u2011integrated AI experimentation lab for engineering (requires PyCharm for full workflow)",
"version": "0.8.17",
"project_urls": {
"Documentation": "https://thalesgroup.github.io/agilab",
"Issues": "https://github.com/ThalesGroup/agilab/issues",
"Source": "https://github.com/ThalesGroup/agilab"
},
"split_keywords": [
"jupyter",
" mlflow",
" asyncio",
" dask",
" rapids",
" streamlit",
" distributed",
" cython",
" cluster",
" dataframe",
" dataset",
" load balancing",
" gen ai",
" copilot",
" agi",
" pycharm",
" datascience"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "3d15ebfb53adaeba7a849ca98ef547bddcde648fa27ab485c38c754c79626751",
"md5": "0c7e546da51a9519a088d9f411fa6c62",
"sha256": "a78497ebc3aadcfee77e1874d70c58c10d495dc751c1083a08abf597317735fb"
},
"downloads": -1,
"filename": "agilab-0.8.17-py3-none-any.whl",
"has_sig": false,
"md5_digest": "0c7e546da51a9519a088d9f411fa6c62",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.11",
"size": 775117,
"upload_time": "2025-10-07T17:32:47",
"upload_time_iso_8601": "2025-10-07T17:32:47.309304Z",
"url": "https://files.pythonhosted.org/packages/3d/15/ebfb53adaeba7a849ca98ef547bddcde648fa27ab485c38c754c79626751/agilab-0.8.17-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-10-07 17:32:47",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "ThalesGroup",
"github_project": "agilab",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "agilab"
}
JSON
