Update Project

.editorconfig

@@ -0,0 +1,26 @@
+root = true
+
+[*]
+charset = utf-8
+end_of_line = lf
+insert_final_newline = true
+trim_trailing_whitespace = true
+indent_style = space
+indent_size = 4
+
+[*.py]
+max_line_length = 120
+
+[*.{yml,yaml,json}]
+indent_size = 2
+
+[*.md]
+indent_size = 2
+# two trailing spaces are a hard line break in Markdown
+trim_trailing_whitespace = false
+
+[*.rst]
+indent_size = 3
+
+[Makefile]
+indent_style = tab

.github/ISSUE_TEMPLATE/bug_report.md

@@ -0,0 +1,30 @@
+---
+name: Bug report
+about: Something does not work as documented
+labels: bug
+---
+
+**What happened**
+A clear description of the bug and what you expected instead.
+
+**Minimal reproduction**
+```python
+# the smallest script that reproduces it
+import lithops
+
+fexec = lithops.FunctionExecutor()
+```
+
+**Traceback / logs**
+```
+paste the full traceback; if possible, run with log_level: DEBUG
+```
+
+**Environment**
+- Lithops version (`lithops --version`):
+- Compute backend (e.g. `localhost`, `aws_lambda`, `code_engine`, `aws_ec2`):
+- Storage backend (e.g. `localhost`, `aws_s3`, `ibm_cos`):
+- Runtime (default or custom image):
+- Python version and OS of the client:
+
+<!-- Remove credentials and account IDs from any configuration you paste. -->

.github/ISSUE_TEMPLATE/config.yml

@@ -0,0 +1,8 @@
+blank_issues_enabled: true
+contact_links:
+  - name: Questions and discussions
+    url: https://github.com/lithops-cloud/lithops/discussions
+    about: Usage questions, ideas and general discussion
+  - name: Security vulnerabilities
+    url: https://github.com/lithops-cloud/lithops/security/advisories/new
+    about: Report security issues privately, not as public issues (see SECURITY.md)

.github/ISSUE_TEMPLATE/feature_request.md

@@ -0,0 +1,17 @@
+---
+name: Feature request
+about: Propose a new feature, backend or API change
+labels: enhancement
+---
+
+**Use case**
+What are you trying to run (workload, scale, compute / storage backend)?
+
+**Problem**
+What is slow, expensive or impossible today?
+
+**Proposal**
+The API or behaviour you would like.
+
+**Alternatives**
+Workarounds you use today.

.github/PULL_REQUEST_TEMPLATE → .github/PULL_REQUEST_TEMPLATE.md

@@ -1,6 +1,21 @@
+## What and why
 
+<!-- What does this change and why is it needed? Link the issue if there is one. -->
+
+## How it was tested
+
+- [ ] `ruff check .`
+- [ ] `pytest -v --backend localhost --storage localhost` (in `lithops/tests`)
+- [ ] On a cloud backend: <!-- e.g. aws_lambda + aws_s3 -->
+
+## Checklist
+
+- [ ] Tests cover the new behaviour / the fixed bug (regression test)
+- [ ] Docs updated (`docs/`, `README.md`, `config/config_template.yaml` for new config keys)
+- [ ] `CHANGELOG.md` entry under the development version
+
+---
 
-
 Developer's Certificate of Origin 1.1
 
        By making a contribution to this project, I certify that:
@@ -25,4 +40,3 @@ Developer's Certificate of Origin 1.1
            personal information I submit with it, including my sign-off) is
            maintained indefinitely and may be redistributed consistent with
            this project or the open source license(s) involved.
-

.github/workflows/python-linting.yml

@@ -5,7 +5,7 @@ on:
     branches:
       - master
     paths:
-      - 'setup.py'
+      - 'pyproject.toml'
       - 'lithops/**'
       - '.github/workflows/python-linting.yml'
 
@@ -18,7 +18,7 @@ concurrency:
 
 jobs:
 
-  flake8:
+  ruff:
     runs-on: ubuntu-latest
     timeout-minutes: 10
 
@@ -34,8 +34,8 @@ jobs:
       - name: Install dependencies
         run: |
           python3 -m pip install --upgrade pip
-          pip3 install -U flake8
+          pip3 install ruff==0.16.8  # same version as .pre-commit-config.yaml
 
-      - name: Lint with flake8
+      - name: Lint with ruff
         run: |
-          flake8 lithops --count --statistics
+          ruff check --output-format=github .

.github/workflows/tests-all-os.yml

@@ -25,7 +25,7 @@ jobs:
         with:
           python-version: ${{ matrix.python-version }}
           cache: 'pip'
-          cache-dependency-path: setup.py
+          cache-dependency-path: pyproject.toml
 
       - name: Install Redis
         # There is no Redis build for Windows. The tests that need a server

.github/workflows/tests.yml

@@ -5,7 +5,7 @@ on:
     branches:
       - master
     paths:
-      - 'setup.py'
+      - 'pyproject.toml'
       - 'lithops/**'
       - '.github/workflows/tests.yml'
 
@@ -48,7 +48,7 @@ jobs:
         with:
           python-version: ${{ matrix.python-version }}
           cache: 'pip'
-          cache-dependency-path: setup.py
+          cache-dependency-path: pyproject.toml
 
       - name: Install Lithops
         run: |

.gitignore

@@ -10,6 +10,10 @@ lithops_*.zip
 *.txt
 *.csv
 *.coverage*
+.pytest_cache/
+
+# Local Lithops config (may hold credentials)
+.lithops_config
 
 # Virtual environments
 .env

.pre-commit-config.yaml

@@ -0,0 +1,12 @@
+repos:
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v6.0.0
+    hooks:
+      - id: check-yaml
+      - id: check-merge-conflict
+      - id: check-added-large-files
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    rev: v0.16.8
+    hooks:
+      # configured in pyproject.toml, same as the Python Linting workflow
+      - id: ruff-check

AGENTS.md

@@ -0,0 +1,112 @@
+# AGENTS.md
+
+Instructions for AI coding agents working in this repository (OpenAI Codex, GitHub Copilot
+coding agent, Cursor, Gemini CLI, Aider, Claude Code and others). Human contributors should
+read [CONTRIBUTING.md](CONTRIBUTING.md); this file summarizes the same rules plus the
+architectural context an agent needs.
+
+## Project overview
+
+Lithops is a Python framework that runs unmodified Python functions at scale on serverless
+platforms, container and batch services, virtual machines and the local machine, behind one
+small API (`FunctionExecutor.call_async()`, `map()`, `map_reduce()`, `wait()`,
+`get_result()`, `Storage`). The architecture is described in
+[docs/source/design.rst](docs/source/design.rst); read it before changing core code.
+
+- **Client:** where `import lithops` runs. Builds jobs, serializes code and data, invokes the
+  workers and collects results. All orchestration lives here.
+- **Compute backend:** where the workers run, one call each.
+- **Object storage:** the only communication bus between client and workers. The client
+  writes the pickled function and data; workers read their input and write their output and
+  status back. There are no direct client-worker connections.
+
+## Setup and commands
+
+```bash
+pip3 install -e '.[tests,dev]'             # editable install + test deps + ruff / pre-commit
+                                           # (backend extras: '.[aws]', '.[gcp]', '.[all]', ...)
+ruff check .                               # lint; must pass (config in pyproject.toml, line length 120)
+
+cd lithops/tests
+pytest -v --timeout=120 --timeout-method=thread --backend localhost --storage localhost
+pytest -v --backend localhost --storage localhost -k test_map    # a subset
+pytest --collect-only                                            # list the tests
+```
+
+**Always pass `--backend localhost --storage localhost`.** Without them the test session
+loads the developer's own configuration (`~/.lithops/config`, `.lithops_config`,
+`LITHOPS_CONFIG_FILE`) and runs against whatever cloud account it points to. Some tests need a
+Redis server on `localhost:6379` and skip themselves when none is reachable.
+
+Documentation is built with Sphinx from `docs/` (`make html`, see [docs/README.md](docs/README.md)).
+
+## Repository map
+
+| Path | Content |
+|---|---|
+| `lithops/__init__.py` | Public API surface (`FunctionExecutor`, `Storage`, `wait`, `get_result`, ...) |
+| `lithops/executors.py` | `FunctionExecutor` and `LocalhostExecutor` / `ServerlessExecutor` / `StandaloneExecutor` |
+| `lithops/config.py` | Loads and merges configuration (dict, env vars, YAML), validates backends |
+| `lithops/job/` | Job creation, function/module serialization (`cloudpickle`), data partitioning |
+| `lithops/invokers.py` | `FaaSInvoker` (per-call invocations) and `BatchInvoker` (one submission, many tasks) |
+| `lithops/future.py`, `wait.py`, `retries.py` | `ResponseFuture`, `wait()` / `get_result()`, `RetryingFunctionExecutor` |
+| `lithops/worker/` | Code that runs on the compute backend: `handler.py` (`function_handler`) and `jobrunner.py` (`JobRunner`) |
+| `lithops/serverless/backends/<name>/` | Serverless / batch compute backends (`<name>.py`, `config.py`, `entry_point.py`) |
+| `lithops/standalone/` | Standalone mode: master / worker VMs coordinated through Redis; `backends/<name>/` for VM providers |
+| `lithops/localhost/` | Localhost compute backend (`v1`, `v2`) |
+| `lithops/storage/` | `Storage` / `InternalStorage`, cloud file APIs; `backends/<name>/` for object stores |
+| `lithops/monitoring/` | Job monitor and its pluggable backends (storage polling, RabbitMQ, Redis, queues) |
+| `lithops/telemetry/` | Prometheus / OpenTelemetry metrics (off by default) |
+| `lithops/multiprocessing/`, `concurrent/`, `util/joblib/` | Drop-in `multiprocessing`, `concurrent.futures` and joblib APIs |
+| `lithops/scripts/` | `lithops` CLI (`cli.py`) and the temporary-data cleaner |
+| `lithops/tests/` | pytest suite (`conftest.py` defines `--backend`, `--storage`, `--config`, `--region`) |
+| `runtime/<backend>/` | Dockerfiles and instructions to build runtimes for each backend |
+| `config/` | `config_template.yaml` (every configuration key) and the configuration guide |
+| `docs/` | Sphinx documentation; `docs/source/compute_config/` and `storage_config/` per backend |
+| `examples/` | Single-file usage examples |
+
+## Architectural invariants
+
+1. **Storage is the bus.** Workers receive their input and return their results and status
+   through object storage (`func_key`, `agg_data`, `output_key`, `status_key`), never through
+   a direct connection to the client.
+2. **Every backend looks the same to the core.** Backend-specific code stays in its backend
+   package; the executor, invoker, job and worker code must not special-case a backend by
+   name when an interface method or a config value can express it.
+3. **The worker must tolerate the environment it runs in.** Worker code runs in minimal
+   runtimes (Lambda, containers, VMs, macOS/Windows localhost). Imports of optional
+   dependencies stay lazy, and a failure in the user's function is reported in its status
+   rather than crashing the worker.
+4. **Configuration keys are documented where users look for them.** A new or changed key goes
+   into `config/config_template.yaml` and the backend's page in `docs/source/compute_config/`
+   or `docs/source/storage_config/` (general `lithops` keys also in
+   `docs/source/lithops_config_keys.csv`).
+5. **Public API compatibility.** `FunctionExecutor`, `Storage`, futures and the
+   `multiprocessing` / `concurrent.futures` layers are used by existing programs; do not
+   change signatures or defaults without a deprecation path and a CHANGELOG entry.
+
+## Conventions
+
+- Python 3.10 - 3.14 (the versions CI tests). Match the style of the surrounding code;
+  `ruff check .` clean with line length 120. Do not run `ruff format`: the code base is not
+  formatted with it and it would rewrite almost every file.
+- Package metadata, dependencies and extras live in `pyproject.toml`. When adding a
+  dependency to an extra, also add it to the `all` extra.
+- Every bug fix includes a regression test; every feature includes tests of its behaviour.
+  Tests must run on the localhost backend and storage; backend-specific code that cannot be
+  exercised locally is tested with fakes or mocks.
+- User-visible changes go into `CHANGELOG.md` under the topmost (development) version
+  heading, in the *Added / Changed / Fixed / Removed* section, prefixed with the component,
+  e.g. `- [Monitoring] ...`. Update `docs/` and `README.md` when behaviour changes.
+- `CONTRIBUTING.md` and `docs/source/contributing.rst` carry the same content: change both.
+- Pull requests target `master`. Keep changes small and focused; do not refactor unrelated
+  code.
+
+## Safety
+
+- Do not run tests, examples or `lithops` CLI commands against real cloud backends (AWS, GCP,
+  Azure, IBM Cloud, Aliyun, Oracle, Kubernetes clusters) and do not build or push runtime
+  images without the maintainer's explicit approval: it creates billable resources and uses
+  real credentials.
+- Never commit credentials or configuration files (`~/.lithops/config`, `.lithops_config`).
+- Do not commit or push unless asked to.

CHANGELOG.md

@@ -24,6 +24,8 @@
 - [Storage] `CloudFileProxy.walk()` matches `os.walk` on a missing path; `cloud_open()` rejects an unsupported mode.
 - [CLI] `job list`, `worker list`, `image delete` and `image list` reject unknown flags.
 - [Joblib] `lithops_args` applied to the batch pool; upload/download pools capped at 32 threads.
+- [Packaging] Package metadata moved from `setup.py` / `setup.cfg` to `pyproject.toml`; `python_requires` raised to `>=3.10`, matching the tested versions.
+- [Development] Linting moved from flake8 to ruff (`ruff check .`, line length 120); new `dev` extra, pre-commit hooks, `.editorconfig`, `AGENTS.md`, `SECURITY.md` and issue / PR templates.
 
 ### Fixed
 - [Core] Serialization, executor IDs, `FuturesList`, module inspection, `chunksize=0`, packaging, ports, and Ctrl+C/`sys.exit()`.

CLAUDE.md

@@ -0,0 +1,18 @@
+# CLAUDE.md
+
+Claude Code reads this file automatically. The project instructions shared by all AI coding
+agents live in [AGENTS.md](AGENTS.md) and are imported here, so there is a single source of
+truth:
+
+@AGENTS.md
+
+## Claude Code specifics
+
+- Use the Python interpreter that has this checkout installed in editable mode
+  (`pip show lithops` shows the *Editable project location*).
+- Run `ruff check .` and the localhost test suite (with `--backend localhost --storage
+  localhost`) before reporting a change as done.
+- Long-running jobs (full test suite, Docker / runtime builds, docs builds) should run in the
+  background.
+- Ask before any action that uses cloud accounts, builds or pushes runtime images, or
+  publishes content.