docs: add API documentation infrastructure

LeRobot's documentation build passes `--not_python_module`, which tells
doc-builder there is no importable Python package and disables `[[autodoc]]`
entirely. The result is that all 90+ pages are hand-written guides and there is
no generated API reference at all.

This is the machinery to change that. It deliberately contains no docstring
changes of its own — every docstring edit lives in the follow-up PR, so this
one can be reviewed as tooling and configuration alone.

**The standard.** `docs/source/writing_docstrings.mdx` is the contract: Google
section headers with Hugging Face type formatting, the machine-checked argument
line, `**Attributes**:`, doc-builder cross-references, fenced doctest examples.
It also records three behaviours that are not discoverable from the source and
were verified against a local build: `[[autodoc]]` silently skips members with
no docstring; doc-builder does not inherit docstrings from base classes, so a
registered config shim whose body is `pass` renders every field with no
description; and module-level aliases resolve to the canonical class.

**Autodoc turned on**, with two changes that are not obvious:

- `--version main` on the main-docs job. Without `--not_python_module`,
  doc-builder resolves the version from `lerobot.__version__` and only maps it
  to the default branch when it contains "dev". transformers relies on that;
  our main carries 0.6.2. Verified by building both ways — dropping the flag
  alone would publish the main docs to /lerobot/v0.6.2/ instead of
  /lerobot/main/ and disable notebook building.
- `pre_command` on both jobs. doc-builder ships a mock-deps registry entry for
  lerobot, so the reusable workflow takes its light-install path, which cannot
  import the package. The heavy dependencies cannot be mocked either: draccus
  runs `register_subclass` at import time and `processor/converters.py` calls
  `functools.singledispatch.register(torch.Tensor)`, which needs a real class.
  `[dataset]` is the only extra required.

Workflow triggers gain `src/**`, since the reference is now generated from
docstrings. `docs/source/api/` is excluded from the prettier hook, which reads
`[[autodoc]]` member lists as lazy paragraph continuations and joins a ten-entry
list onto one line.

Nine API reference pages, scaffolded with each module's base class.

**Doctests.** `LeRobotDocTestParser` is mandatory rather than optional here:
ruff's `docstring-code-format = true` drops the blank line before a closing
fence, after which stdlib's `_EXAMPLE_RE` reads the fence as expected output and
every example with output fails. It is written against the installed pytest
rather than copied from transformers, whose version predates pytest 9's
`import_path` signature and its own fix for the `@property` line-number bug.
`preprocess_string` also diverges: the upstream fenced-block split puts a
single-line example's code in a chunk with no `>>>` in it, so neither the CUDA
skip nor the `+IGNORE_RESULT` injection fires for it.

**Checkers.** `utils/check_docstrings.py` is the ~300-line core of the
2203-line transformers original; the `@auto_docstring` system, modular
propagation, GitPython and `checkers.py` are not ported.
`utils/check_config_docstrings.py` checks that every registered robot config
documents its port and calibration semantics.

**Gates**, all set to values that pass today: ruff `D` with per-file-ignores
per unconverted module, `interrogate` at `fail-under = 52` against a measured
52.1%, and Makefile targets wired into the quality workflow. The doctest
allowlist ships empty and the `doctest` target handles that, because the files
carrying runnable examples arrive with the docstring PR.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

.github/workflows/documentation.yml

@@ -24,19 +24,24 @@ on:
         required: false
         type: string
 
-  # Triggers the workflow on push events to main for the docs folder
+  # Triggers on pushes to main that touch the docs or the sources the API reference is generated from.
+  # `src/**` is included because the API reference is built from docstrings via `[[autodoc]]`: without it,
+  # published API pages would go stale as soon as a docstring changed.
   push:
     branches:
       - main
     paths:
       - "docs/**"
+      - "src/**"
 
-  # Triggers the workflow on pull request events targeting main for the docs folder
+  # Same for pull requests, so a docstring change gets a preview build and a broken `[[autodoc]]` path
+  # fails the PR rather than main.
   pull_request:
     branches:
       - main
     paths:
       - "docs/**"
+      - "src/**"
 
   release:
     types: [published]
@@ -59,12 +64,21 @@ jobs:
     with:
       commit_sha: ${{ github.sha }}
       package: lerobot
+      # doc-builder ships a mock-deps registry entry for lerobot, so the reusable workflow takes its
+      # "light install" path: `pip install ./lerobot --no-deps` plus a handful of real dependencies.
+      # That is not enough to import lerobot — draccus runs `register_subclass` at import time and
+      # `processor/converters.py` calls `functools.singledispatch.register(torch.Tensor)`, neither of
+      # which works against a mock. Install the package for real before the build.
+      pre_command: uv pip install "./lerobot[dataset]"
+      # `--version main` is load-bearing: without `--not_python_module`, doc-builder falls back to
+      # `lerobot.__version__` and only maps that to the default branch when it contains "dev". Our main
+      # branch carries a release version (0.6.2), so omitting this would publish the main docs to
+      # /lerobot/v0.6.2/ instead of /lerobot/main/ and disable notebook building.
       additional_args: >-
-        --not_python_module
         ${{
           (github.event_name == 'release' && format('--version {0}', github.event.release.tag_name)) ||
           (inputs.version != '' && format('--version {0}', inputs.version)) ||
-          ''
+          '--version main'
         }}
     secrets:
       token: ${{ secrets.HUGGINGFACE_PUSH }}
@@ -83,4 +97,6 @@ jobs:
       commit_sha: ${{ github.event.pull_request.head.sha }}
       pr_number: ${{ github.event.number }}
       package: lerobot
-      additional_args: --not_python_module
+      # See the comment on build_main_docs. The PR workflow passes its own `--version pr_<n>`, so no
+      # additional_args are needed here.
+      pre_command: uv pip install "./lerobot[dataset]"

.github/workflows/quality.yml

@@ -56,3 +56,41 @@ jobs:
         uses: pre-commit/action@2c7b3805fd2a0fd8c1884dcaebf91fc102a13ecd  # v3.0.1
         with:
           extra_args: --all-files --show-diff-on-failure --color=always
+
+  # This job runs the examples in our docstrings and validates the doctest allowlist.
+  # See docs/source/writing_docstrings.mdx for the standard these enforce.
+  doc-checks:
+    name: Run Documentation Checks (Doctests)
+    runs-on: ubuntu-latest
+    env:
+      # Examples that need a physical robot, a serial port or a Hub download are skipped by content.
+      # Everything else has to actually run. See src/lerobot/utils/doctest_utils.py.
+      SKIP_HARDWARE_DOCTEST: "1"
+      SKIP_CUDA_DOCTEST: "1"
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1  # v7.0.1
+        with:
+          persist-credentials: false
+
+      - name: Setup uv and Python
+        uses: astral-sh/setup-uv@11f9893b081a58869d3b5fccaea48c9e9e46f990  # v8.3.2
+        with:
+          enable-cache: true
+          version: "0.11.30"
+          python-version: "3.12"
+
+      - name: Install dependencies
+        run: uv sync --locked --extra test --extra dataset
+
+      - name: Check the doctest list is sorted and its paths exist
+        run: make check-doctest-list
+
+      - name: Check documented arguments match their signatures
+        run: make check-docstrings
+
+      - name: Check docstring coverage has not regressed
+        run: uv run --with interrogate interrogate --config=pyproject.toml
+
+      - name: Run doctests
+        run: make doctest