docs-anonymizer 0.2.2__tar.gz → 0.2.4__tar.gz

{docs_anonymizer-0.2.2 → docs_anonymizer-0.2.4}/CHANGELOG.md

@@ -11,6 +11,42 @@ static release metadata, and posts the release notification.
 
 ## [Unreleased]
 
+## [0.2.4] - 2026-05-19
+
+### Added
+
+- Vendor-neutral agent workflow infrastructure under ``.agents/skills`` so
+  Claude Code, Codex, and generic agents share one canonical workflow source.
+  ``.claude/skills`` entries are now compatibility shims that delegate to the
+  canonical files, with adapter notes for Claude, Codex, and generic agents.
+
+### Fixed
+
+- Agent workflow consistency tests now validate YAML frontmatter with
+  ``yaml.safe_load``. This prevents invalid multiline ``description`` fields
+  from silently breaking skill discovery while still keeping Claude shims and
+  canonical ``.agents`` skills in sync.
+
+## [0.2.3] - 2026-05-19
+
+### Fixed
+
+- Add ``setuptools<81`` as a hard runtime dependency in
+  ``[project] dependencies``. Without the runtime pin, ``uv tool
+  install docs-anonymizer`` end-user environments resolved the latest
+  setuptools (>=81), which dropped the bundled ``pkg_resources``
+  package; the first RU NER call then crashed with::
+
+      File ".../pymorphy2/analyzer.py", line 114, in _iter_entry_points
+          import pkg_resources
+      ModuleNotFoundError: No module named 'pkg_resources'
+
+  Surfaced in the pilot run on the first real RU ``.docx`` document.
+  The existing ``[tool.uv] constraint-dependencies`` line was a
+  uv-only override and did not propagate into the published wheel
+  metadata. Lifting this pin requires upgrading natasha to a release
+  that uses pymorphy3.
+
 ## [0.2.2] - 2026-05-19
 
 ### Added

{docs_anonymizer-0.2.2 → docs_anonymizer-0.2.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: docs-anonymizer
-Version: 0.2.2
+Version: 0.2.4
 Summary: Offline document anonymizer for legal teams
 Project-URL: Homepage, https://anonymizer.site
 License-Expression: AGPL-3.0
@@ -32,6 +32,7 @@ Requires-Dist: pymupdf>=1.24
 Requires-Dist: python-docx>=1.1
 Requires-Dist: python-multipart>=0.0.6
 Requires-Dist: pyyaml>=6.0
+Requires-Dist: setuptools<81
 Requires-Dist: spacy>=3.7
 Requires-Dist: sse-starlette>=2.0
 Requires-Dist: transliterate>=1.10

{docs_anonymizer-0.2.2 → docs_anonymizer-0.2.4}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "docs-anonymizer"
-version = "0.2.2"
+version = "0.2.4"
 description = "Offline document anonymizer for legal teams"
 readme = "README.md"
 # Upper bound: ``blis 0.7.11`` (pulled in transitively via ``spacy 3.7`` →
@@ -54,6 +54,14 @@ dependencies = [
     "Babel>=2.14",
     "httpx>=0.27",
     "packaging>=24.0",
+    # ``natasha.MorphVocab`` initialises ``pymorphy2`` which calls
+    # ``import pkg_resources`` at runtime. Setuptools >= 81 dropped the
+    # bundled ``pkg_resources`` package, so the import crashes the first
+    # RU NER call. We need a strict runtime cap (not a uv-only constraint)
+    # so end-user ``uv tool install`` environments pick a compatible
+    # setuptools out of the box. Lifting this pin requires upgrading
+    # natasha to a release that uses pymorphy3.
+    "setuptools<81",
 ]
 
 [project.scripts]
@@ -73,10 +81,10 @@ build-backend = "hatchling.build"
 # ``thinc.backends.numpy_ops``). Lifting this pin requires upgrading to
 # spaCy 3.8+, which falls outside the locked ``en_core_web_lg-3.7.1`` model.
 #
-# Natasha's ``MorphVocab`` initialises ``pymorphy2`` (not pymorphy3), which
-# calls ``import pkg_resources`` at runtime. Setuptools >= 81 dropped the
-# bundled ``pkg_resources`` package, so we constrain setuptools below that
-# version until pymorphy2 is patched or replaced upstream.
+# ``setuptools<81`` is also declared as a runtime dependency in
+# ``[project] dependencies`` so end-user ``uv tool install`` picks it up.
+# It remains here as a constraint to lock the dev-install path through
+# the same path. See the inline comment next to ``setuptools<81`` above.
 constraint-dependencies = ["numpy<2", "setuptools<81"]
 
 [tool.uv.sources]

docs_anonymizer-0.2.4/tests/unit/test_agent_workflows.py

@@ -0,0 +1,92 @@
+"""Consistency checks for vendor-neutral agent workflows."""
+
+from __future__ import annotations
+
+from pathlib import Path
+
+import yaml
+
+ROOT = Path(__file__).resolve().parents[2]
+
+EXPECTED_SKILLS = {
+    "analyze-feedback",
+    "annotate-corpus",
+    "generate-synthetic-doc",
+    "improve-detector",
+    "regression-check",
+    "release-patch",
+    "reproduce-failure",
+}
+
+
+def _skill_names(base: Path) -> set[str]:
+    return {path.parent.name for path in base.glob("*/SKILL.md")}
+
+
+def _frontmatter_metadata(text: str) -> dict[str, object]:
+    frontmatter = text.split("---\n", 2)[1]
+    parsed = yaml.safe_load(frontmatter)
+    assert isinstance(parsed, dict)
+    return parsed
+
+
+def _frontmatter_description(text: str) -> str:
+    description = _frontmatter_metadata(text).get("description")
+    assert isinstance(description, str)
+    return description
+
+
+def test_canonical_agent_skills_exist_and_match_frontmatter() -> None:
+    skills_dir = ROOT / ".agents" / "skills"
+
+    assert _skill_names(skills_dir) == EXPECTED_SKILLS
+    for skill_name in EXPECTED_SKILLS:
+        text = (skills_dir / skill_name / "SKILL.md").read_text(encoding="utf-8")
+        assert text.startswith("---\n")
+        metadata = _frontmatter_metadata(text)
+        assert metadata.get("name") == skill_name
+        assert _frontmatter_description(text).startswith("Use when ")
+
+
+def test_claude_skills_are_compatibility_shims() -> None:
+    claude_dir = ROOT / ".claude" / "skills"
+
+    assert _skill_names(claude_dir) == EXPECTED_SKILLS
+    for skill_name in EXPECTED_SKILLS:
+        canonical_text = (ROOT / ".agents" / "skills" / skill_name / "SKILL.md").read_text(
+            encoding="utf-8"
+        )
+        text = (claude_dir / skill_name / "SKILL.md").read_text(encoding="utf-8")
+        metadata = _frontmatter_metadata(text)
+        assert metadata.get("name") == skill_name
+        assert _frontmatter_description(text).startswith("Use when ")
+        assert _frontmatter_description(text) == _frontmatter_description(canonical_text)
+        assert "Claude Code" in text
+        assert "compatibility shim" in text.casefold()
+        assert f".agents/skills/{skill_name}/SKILL.md" in text
+        assert "source of truth" in text
+
+
+def test_agent_docs_point_to_canonical_workflows() -> None:
+    agents_md = (ROOT / "AGENTS.md").read_text(encoding="utf-8")
+    guide = (ROOT / ".agents" / "AGENT_GUIDE.md").read_text(encoding="utf-8")
+
+    assert ".agents/skills" in agents_md
+    assert ".claude/skills" in agents_md
+    assert "compatibility" in agents_md.casefold()
+    for skill_name in EXPECTED_SKILLS:
+        assert guide.count(f".agents/skills/{skill_name}/SKILL.md") >= 1
+
+
+def test_codex_adapter_documents_manual_skill_loading() -> None:
+    adapter = (ROOT / ".agents" / "adapters" / "codex.md").read_text(encoding="utf-8")
+
+    assert "does not auto-discover skills" in adapter
+
+
+def test_agent_plan_uses_repository_local_instructions() -> None:
+    plan = (
+        ROOT / "docs" / "superpowers" / "plans" / "2026-05-19-agent-workflows-neutral.md"
+    ).read_text(encoding="utf-8")
+
+    assert "superpowers:" not in plan

docs_anonymizer-0.2.4/tests/unit/test_pyproject_dependency_pins.py

@@ -0,0 +1,150 @@
+"""Guard critical runtime dependency pins in ``pyproject.toml``.
+
+Each pin documented here was added because a real install failure
+escaped pure-pytest coverage — the constraint lives in published wheel
+metadata so end-user ``uv tool install`` environments resolve compatible
+versions. Removing or weakening any of these without a coordinated
+upstream fix has historically broken pilot installs.
+
+These tests are intentionally narrow text-level assertions on the
+declared dependency list; they're not a substitute for an actual
+``uv tool install`` end-to-end smoke run, but they do catch the obvious
+regression of "someone deleted the pin while bumping versions".
+"""
+
+from __future__ import annotations
+
+import tomllib
+from pathlib import Path
+
+import pytest
+from packaging.requirements import Requirement
+from packaging.version import Version
+
+_PYPROJECT = Path(__file__).parents[2] / "pyproject.toml"
+
+
+@pytest.fixture(scope="module")
+def project_metadata() -> dict:
+    return tomllib.loads(_PYPROJECT.read_text(encoding="utf-8"))["project"]
+
+
+@pytest.fixture(scope="module")
+def runtime_dependencies(project_metadata: dict) -> list[Requirement]:
+    return [Requirement(raw) for raw in project_metadata["dependencies"]]
+
+
+def _find_dep(deps: list[Requirement], package_name: str) -> Requirement | None:
+    """Return the parsed ``Requirement`` for ``package_name`` (case-insensitive)."""
+
+    lowered = package_name.lower()
+    for req in deps:
+        if req.name.lower() == lowered:
+            return req
+    return None
+
+
+def test_setuptools_capped_below_81(runtime_dependencies: list[Requirement]) -> None:
+    """``natasha`` → ``pymorphy2`` runtime-imports ``pkg_resources``.
+
+    Setuptools 81 dropped the bundled ``pkg_resources`` package, so the
+    first RU NER call crashes in any end-user environment that resolves
+    setuptools >= 81. The pin was added in 0.2.3 after pilot crash;
+    removing it regresses the pilot. Lift only after natasha publishes
+    a pymorphy3-based release.
+
+    Assert semantically via ``Requirement.specifier.contains`` so typos
+    like ``>=81`` or ``!=81`` don't sneak past a substring check.
+    """
+
+    req = _find_dep(runtime_dependencies, "setuptools")
+    assert req is not None, (
+        "setuptools must be declared as a runtime dependency with an "
+        "upper-bound. natasha.MorphVocab → pymorphy2 calls "
+        "import pkg_resources, which setuptools 81+ no longer ships."
+    )
+    # Pre-release versions of 81 (e.g. 81.0.0a1) would still drop pkg_resources;
+    # include them in the exclusion check too.
+    assert not req.specifier.contains("81.0.0", prereleases=True), (
+        f"setuptools 81.0.0 must be excluded by the version specifier; "
+        f"got {req!s}. See CHANGELOG [0.2.3]."
+    )
+    assert not req.specifier.contains("82.0.0", prereleases=True), (
+        f"setuptools 82.0.0 must also be excluded; got {req!s}."
+    )
+    # And the last known-good release must still satisfy the pin.
+    assert req.specifier.contains(Version("80.10.2")), (
+        f"setuptools 80.10.2 (known good) must satisfy the pin; got {req!s}."
+    )
+
+
+def test_en_core_web_lg_runtime_dependency(
+    runtime_dependencies: list[Requirement],
+) -> None:
+    """spaCy English model is required by ``person_en`` / ``company_en``.
+
+    It is not on PyPI proper — the installer scripts inject the URL via
+    ``uv tool install --with "en-core-web-lg @ <github wheel URL>"``.
+    The bare requirement must remain in ``[project] dependencies`` so
+    the resolver validates the link target; the URL itself lives in
+    ``[tool.uv.sources]`` and the installer ``--with`` flag.
+    """
+
+    req = _find_dep(runtime_dependencies, "en-core-web-lg")
+    assert req is not None, (
+        "en-core-web-lg must be declared as a runtime dependency so the "
+        "resolver validates the spaCy model wheel injected at install "
+        "time via uv tool install --with."
+    )
+    # PyPI rejects direct URL references in published metadata
+    # (see CHANGELOG [0.2.0] direct-URL hotfix). The bare requirement
+    # must have neither a URL nor a strict specifier locking the model
+    # to a specific PyPI version.
+    assert req.url is None, (
+        f"en-core-web-lg requirement must not embed a direct URL in "
+        f"published wheel metadata (PyPI rejects it); got {req!s}. The "
+        f"URL lives in [tool.uv.sources] for dev installs and in the "
+        f"installer --with flag for end users."
+    )
+
+
+def test_requires_python_capped_below_313(project_metadata: dict) -> None:
+    """``blis 0.7.11`` only ships pre-built wheels for cp310..cp312.
+
+    On Python 3.13/3.14 uv falls back to building blis from sdist and
+    the build crashes against the CPython 3.13/3.14 C ABI (removed
+    ``_PyDict_SetItem_KnownHash``, new ``_PyLong_AsByteArray`` sig).
+    The upper bound was added in 0.2.1; lifting it requires upgrading
+    to spacy >= 3.8 and refreshing the ``en_core_web_lg`` model URL.
+
+    Uses ``SpecifierSet.contains`` so a typo like ``>=3.13`` would still
+    fail the test instead of slipping past a substring check.
+    """
+
+    from packaging.specifiers import SpecifierSet
+
+    requires_python = SpecifierSet(project_metadata["requires-python"])
+    assert not requires_python.contains("3.13.0", prereleases=True), (
+        f"requires-python must exclude 3.13: got {requires_python!s}. "
+        "blis 0.7.11 has no cp313/cp314 wheels. See CHANGELOG [0.2.1]."
+    )
+    assert not requires_python.contains("3.14.0", prereleases=True), (
+        f"requires-python must exclude 3.14: got {requires_python!s}."
+    )
+    assert requires_python.contains("3.12.0"), (
+        f"requires-python must still allow 3.12: got {requires_python!s}."
+    )
+
+
+def test_distribution_name_is_docs_anonymizer(project_metadata: dict) -> None:
+    """The PyPI distribution name is ``docs-anonymizer``, not ``anonymizer``.
+
+    The bare ``anonymizer`` name was unavailable on PyPI; ``docs-anonymizer``
+    was reserved in 0.0.1 and shipped functionally in 0.2.0. The Python
+    import name is unrelated and stays ``anonymizer``.
+    """
+
+    assert project_metadata["name"] == "docs-anonymizer", (
+        "distribution name on PyPI must stay 'docs-anonymizer'. "
+        "Python import name 'anonymizer' is decoupled from this."
+    )