testsAndMisc/.pre-commit-config.yaml
Krzysztof kuhy Rudnicki ec2da0b575
Some checks are pending
dependency freshness / dependencies on newest stable (push) Waiting to run
md-naming / markdown naming convention (push) Waiting to run
Pre-commit checks / pre-commit (push) Waiting to run
Python tests / test (push) Waiting to run
Shell tests / shell-tests (push) Waiting to run
Shell tests / arch-tests (push) Waiting to run
helium: unlock the keyring through secretstorage; anchor the running-browser check
`secret-tool lookup` is non-interactive — on a locked keyring it returns
nothing and shows no dialog — and a busctl Unlock+Prompt pair cannot work
either, because Secret Service prompt objects die with the D-Bus
connection that created them. secretstorage keeps one connection:
get_default_collection().unlock() pops the gcr dialog and blocks until it
is answered. Verified on the real profile: 5 v11 rows -> v10, then with
the keyring re-locked Helium (through the wrapper, so the flags file and
LeechBlock apply) lists all 5 logins and keeps every row.

`pgrep -f /opt/helium-browser-bin` also matched the invoking shell when
its argv mentioned the path, so the guard refused on an idle machine;
now anchored to `^/opt/helium-browser-bin/`.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01M8YGXVZdNUG79q5DGsdMm3
2026-09-20 18:29:49 +02:00

625 lines
27 KiB
YAML

# ==============================================================================
# Pre-commit Configuration - Multi-language Linting & Formatting
# ==============================================================================
# Install: pre-commit install && pre-commit install --hook-type pre-push
# Fast lint: pre-commit run --all-files (linters only, ~10 s)
# Full suite: pre-commit run --all-files --hook-stage pre-push (+ tests)
# Update hooks: pre-commit autoupdate
# ==============================================================================
# Global settings
default_language_version:
python: python3
# By default every hook runs ONLY at commit-time. Pre-push runs only the
# hooks that explicitly opt in via `stages: [pre-push]` (currently
# pytest-coverage + prettier). This keeps `git push` cheap; commit-time
# already enforces the full lint suite.
default_stages: [pre-commit]
# Fail fast on first error (set to false to see all errors)
fail_fast: false
# Configuration
ci:
autofix_commit_msg: "style: auto-fix by pre-commit hooks"
autoupdate_commit_msg: "chore: update pre-commit hooks"
repos:
# ===========================================================================
# GENERAL HOOKS - File formatting and validation
# ===========================================================================
- repo: https://github.com/pre-commit/pre-commit-hooks
rev: v4.6.0
hooks:
- id: trailing-whitespace
args: [--markdown-linebreak-ext=md]
- id: end-of-file-fixer
- id: check-yaml
args: [--unsafe]
- id: check-json
# Exclude JSONC files (VS Code configs, TypeScript configs)
exclude: ^(\.vscode/|.*/\.vscode/|.*tsconfig.*\.json)
- id: check-toml
- id: check-xml
- id: check-added-large-files
args: [--maxkb=2000]
- id: check-merge-conflict
- id: check-case-conflict
- id: check-symlinks
- id: check-executables-have-shebangs
- id: check-shebang-scripts-are-executable
- id: detect-private-key
- id: debug-statements
- id: name-tests-test
args: [--pytest-test-first]
- id: check-ast
- id: check-builtin-literals
- id: check-docstring-first
- id: fix-byte-order-marker
- id: mixed-line-ending
args: [--fix=lf]
- id: requirements-txt-fixer
# ===========================================================================
# BINARY BLOCKER - Prevent binary/image files from being committed
# ===========================================================================
- repo: local
hooks:
- id: no-binaries
name: Block binary/image files
entry: meta/scripts/check_no_binaries.sh
language: script
always_run: false
- id: file-length-cap
name: Enforce the 250-line file cap
# Repo-wide (--all), not per-file: a file that grows past the cap
# without being staged in this commit must still be caught. Wired only
# after the campaign that brought every file under it — turning this on
# earlier would have blocked every unrelated commit in the repo.
entry: bash meta/scripts/check_file_length.sh --all
language: system
pass_filenames: false
always_run: true
- id: directory-depth-cap
name: Enforce the directory-depth cap
# Repo-wide and ENFORCING. It ran --warn while the 46 paths under
# periodic_background/{hosts,digital_wellbeing} were over the cap;
# both units are now their own repos (kuhyx/hosts-blocker,
# kuhyx/digital-wellbeing) and the count is 0, so the cap is real.
# Re-adding --warn would silently re-open the hole it was there to
# close: put the offending path somewhere legal instead.
entry: bash meta/scripts/check_directory_depth.sh --all
language: system
pass_filenames: false
always_run: true
- id: repo-size-cap
name: Enforce the source-file ceiling
# Counts tracked files only, excluding vendored and hook-generated
# trees. Prints the count on every run so the ceiling is auditable.
entry: bash meta/scripts/check_repo_size.sh
language: system
pass_filenames: false
always_run: true
- id: ci-baseline-green
name: Refuse to build on a red CI baseline
# Checks the last COMPLETED run of each required workflow on main.
# Fails closed on a definitively red baseline; warns and passes on
# anything it cannot determine (offline, no gh, no runs yet), because
# the real per-change guarantee is the pre-push ci-mirror hook and
# blocking every commit on network reachability is not worth it.
# Bypass a deliberate fix-forward commit with CI_GREEN_SKIP=1.
entry: bash meta/scripts/check_ci_green.sh
language: system
pass_filenames: false
always_run: true
- id: ai-evidence-contract
name: Require AI evidence artifacts for code changes
entry: meta/scripts/check_ai_evidence.sh
language: script
pass_filenames: false
always_run: true
- id: ai-multifile-contract
name: Require workflow contract for multi-file code changes
entry: meta/scripts/check_agent_contract.sh
language: script
pass_filenames: false
always_run: true
- id: append-only-sessions
name: Enforce append-only session logs
entry: meta/scripts/check_append_only_sessions.sh
language: script
pass_filenames: false
always_run: true
# ===========================================================================
# POLLING SCRIPT LINTER - Detect fork-storm anti-patterns in shell scripts
# ===========================================================================
- repo: local
hooks:
- id: no-inline-python
name: Block Python logic embedded in shell scripts
entry: meta/scripts/check_no_inline_python.sh
language: script
types_or: [shell, zsh]
- id: no-polling-antipatterns
name: Block polling script anti-patterns
entry: meta/scripts/check_polling_antipatterns.sh
language: script
types: [shell]
require_serial: true
exclude: ^(\.git/|tests/)
- id: shell-coverage-ratchet
name: Every shell library has a test suite beside it
# The ratchet from meta/shell-coverage-allowlist.txt: a lib under a
# lib/ directory needs a tests/run_all.sh, unless it is allowlisted.
# Wired here on 2026-08-23. It existed but was never in pre-commit,
# so it had been silently unenforced -- a new untested lib could land
# freely. That was only affordable to fix once 578c4a6c dropped the
# jail-measured 100% bar: --all went from >600s to 0.31s, and the
# per-file mode this hook uses is milliseconds.
entry: bash meta/scripts/check_shell_coverage.sh
language: system
types: [shell]
exclude: ^third_party/
# ===========================================================================
# NOQA BLOCKER - Zero tolerance for noqa/type:ignore suppression comments
# ===========================================================================
- repo: local
hooks:
- id: no-noqa
name: Block noqa comments
entry: '(?i)#\s*(noqa|type:\s*ignore)'
language: pygrep
types: [python]
- id: no-ruff-noqa
name: Block ruff noqa file-level comments
entry: '(?i)#\s*ruff:\s*noqa'
language: pygrep
types: [python]
# ===========================================================================
# RUFF - Fast Python linter and formatter (replaces black, isort, flake8, etc.)
# ===========================================================================
- repo: https://github.com/astral-sh/ruff-pre-commit
rev: v0.15.2
hooks:
# Linter - run first to catch issues
- id: ruff
args:
- --fix
- --unsafe-fixes
- --exit-non-zero-on-fix
- --show-fixes
types_or: [python, pyi]
# Formatter - run after linting
- id: ruff-format
types_or: [python, pyi]
# ===========================================================================
# MYPY - Static type checking (per-commit on changed files only)
# Was on pre-push, but force-push diffs caused full-repo scans + OOM. On
# pre-commit it sees only the file(s) currently staged → near-instant.
# ===========================================================================
- repo: https://github.com/pre-commit/mirrors-mypy
rev: v1.13.0
hooks:
- id: mypy
stages: [pre-commit]
args:
- --ignore-missing-imports
- --no-error-summary
- --disable-error-code=no-untyped-def
- --disable-error-code=no-untyped-call
- --disable-error-code=var-annotated
- --disable-error-code=no-any-unimported
- --disable-error-code=type-arg
- --disable-error-code=no-any-return
- --disable-error-code=misc
- --disable-error-code=unused-ignore
- --disable-error-code=unreachable
- --disable-error-code=assignment
- --disable-error-code=no-redef
- --disable-error-code=attr-defined
- --disable-error-code=arg-type
- --disable-error-code=union-attr
- --disable-error-code=call-overload
- --disable-error-code=return-value
- --disable-error-code=redundant-cast
- --disable-error-code=empty-body
- --disable-error-code=list-item
exclude: >-
(?x)^(
Bash/.*|
\.venv/.*|
linux_configuration/misc/tools/.*
)$
additional_dependencies:
- types-requests
- types-PyYAML
- types-python-dateutil
# ===========================================================================
# PYLINT - Comprehensive Python linter (per-commit on changed files only)
# Was on pre-push, but force-push diffs caused full-repo scans + OOM.
# ===========================================================================
- repo: https://github.com/pylint-dev/pylint
rev: v3.3.2
hooks:
- id: pylint
stages: [pre-commit]
args:
- --rcfile=pyproject.toml
- --fail-under=10
- --jobs=4
additional_dependencies:
- pytest
- python-chess
- requests
- pygame
- pillow
# archwiki_rag parses the local arch-wiki-docs HTML dump. pylint runs
# in its own venv, so a package being in meta/requirements.txt is not
# enough for it to resolve the import.
- beautifulsoup4
- markdownify
# android_ui parses `uiautomator dump` output. Same reason as
# beautifulsoup4 above: requirements.txt is not on pylint's path.
- defusedxml
# artgate computes image gates over pixel arrays. Same reason again:
# it is the first package here to import numpy.
- numpy
- gatelock @ git+https://github.com/kuhyx/utils@gatelock-v0.1.0#subdirectory=gatelock
- anthropic
- rich
- typer
# linux_configuration/installers: verify_helium drives Helium
# over the DevTools websocket; migrate_helium_passwords_to_basic re-encrypts
# Chromium os_crypt blobs (AES-CBC).
- websockets
- cryptography
- secretstorage
# Test suites and conftest fixtures are linted separately; they use
# patterns (protected-access, missing docstrings, fixture shadowing)
# that don't belong in the source-code 10/10 gate.
exclude: ^(Bash/|\.venv/)|(^|/)(tests/|conftest\.py)
# ===========================================================================
# BANDIT - Security linter (per-commit on changed files only)
# Was on pre-push, but force-push diffs caused full-repo scans + OOM.
# ===========================================================================
- repo: https://github.com/PyCQA/bandit
rev: 1.7.10
hooks:
- id: bandit
stages: [pre-commit]
args:
- -c
- pyproject.toml
- --severity-level=high
- --confidence-level=medium
- --skip=B113
additional_dependencies: ["bandit[toml]"]
exclude: ^(Bash/|\.venv/|tests/|.*test.*\.py$)
# ===========================================================================
# PYTEST + COVERAGE - Run tests and enforce 100% code coverage
# Only tests for subpackages with changed files are run (see script).
# Uses pytest-xdist (-n auto) to parallelize across all CPUs.
# ===========================================================================
- repo: local
hooks:
- id: pytest-coverage
name: pytest with coverage enforcement
# Serialised against builds and pacman transactions on this 7.6 GB
# machine via the heavy-job lock. run_with_heavy_lock.sh is a plain
# passthrough on any clone that has not installed one.
entry: bash meta/scripts/run_with_heavy_lock.sh pytest python meta/scripts/pytest_changed_packages.py
language: system
types: [python]
pass_filenames: true
require_serial: true
stages: [pre-commit]
- id: requirements-sync
name: requirements.txt covers all python_pkg imports
entry: python3 meta/scripts/check_requirements_sync.py --package python_pkg
language: system
pass_filenames: false
files: ^(python_pkg/.*\.py|requirements\.txt)$
- id: requirements-test-sync
name: requirements-test.txt covers all python_pkg imports
# python-tests.yml installs this SUBSET rather than the full list, so
# it needs the same completeness guarantee. Without this hook a new
# third-party import would pass locally and fail in CI with a bare
# ModuleNotFoundError.
entry: python3 meta/scripts/check_requirements_sync.py --package python_pkg --requirements meta/requirements-test.txt
language: system
pass_filenames: false
files: ^(python_pkg/.*\.py|meta/requirements-test\.txt)$
- id: session-autopsy-hook-mutations
name: session_autopsy hook tests are non-vacuous (mutation)
entry: python3 meta/scripts/mutate_session_autopsy_hooks.py
language: system
pass_filenames: false
require_serial: true
# Scoped to the two files that can invalidate the proof: the suite
# itself and the mutation definitions. Takes ~40s, so it must not run
# on unrelated commits. Skips itself where ~/.claude is absent (CI).
files: ^(linux_configuration/tests/test_session_autopsy_hooks\.py|meta/scripts/mutate_session_autopsy_hooks\.py)$
- id: ci-mirror
name: CI mirror (clean-venv install + --all-files + changed-pkg pytest)
entry: bash meta/scripts/ci_mirror.sh
language: system
pass_filenames: false
always_run: true
require_serial: true
stages: [pre-push]
# ===========================================================================
# VULTURE - Dead code detection (disabled - doesn't work well with pre-commit)
# ===========================================================================
# - repo: https://github.com/jendrikseipp/vulture
# rev: v2.13
# hooks:
# - id: vulture
# args:
# - --min-confidence=80
# - --exclude=.venv,Bash,__pycache__
# exclude: ^(Bash/|\.venv/)
# ===========================================================================
# PYUPGRADE - Upgrade Python syntax (disabled - incompatible with Python 3.14)
# ===========================================================================
# - repo: https://github.com/asottile/pyupgrade
# rev: v3.19.0
# hooks:
# - id: pyupgrade
# args:
# - --py310-plus
# ===========================================================================
# CODESPELL - Spell checking in code (expanded ignore list for non-English)
# ===========================================================================
- repo: https://github.com/codespell-project/codespell
rev: v2.3.0
hooks:
- id: codespell
args:
- --skip=*.json,*.lock,*-lock.yaml,*.min.js,*.min.css,.git,__pycache__,.venv,*.txt
# Many entries below are Polish words, not typos: they are field
# names from Polish data formats (e.g. `dokument` in the e-paragon /
# JPK e-receipt schema that billsplit parses). "Correcting" them to
# English breaks the parsers that match on them.
# `daa` is MediaTek's Download Agent Authentication, a BROM security
# feature named in the mtk_root notes - not a misspelling of "data".
# `artic` is SearXNG's engine id for the Art Institute of Chicago
# (setup_searxng.sh), not a misspelling of "arctic" - renaming it
# would stop matching the engine.
# `imput` is Helium's vendor (imput.net): its profile dir is
# `~/.config/net.imput.helium` and the installers/*helium* scripts spell it out.
- --ignore-words-list=als,ans,ect,nd,som,sur,te,nam,numer,lew,sie,wil,postion,clen,ther,folow,derrive,ony,tje,noe,theses,crate,doubleclick,wile,tabel,pary,blok,bloc,proces,serwer,parametr,adres,hart,dout,metod,tekst,synonim,grup,mosty,lokal,skalar,milion,nowe,tre,hel,alph,iif,dokument,daa,artic,imput
exclude: ^(Bash/ffmpeg-build/|LaTeX/|.*\.geojson$)
# ===========================================================================
# DOCFORMATTER - Format docstrings (disabled - causes recursion errors)
# ===========================================================================
# - repo: local
# hooks:
# - id: docformatter
# name: docformatter
# entry: docformatter
# language: system
# types: [python]
# args:
# - --in-place
# - --wrap-summaries=88
# - --wrap-descriptions=88
# ===========================================================================
# INTERROGATE - Docstring coverage (disabled - causes recursion on large files)
# ===========================================================================
# - repo: https://github.com/econchick/interrogate
# rev: 1.7.0
# hooks:
# - id: interrogate
# args:
# - --fail-under=0
# - --verbose
# - --ignore-init-method
# - --ignore-init-module
# - --ignore-magic
# - --ignore-private
# - --ignore-semiprivate
# - --exclude=Bash,.venv,__pycache__
# pass_filenames: false
# ===========================================================================
# AUTOFLAKE - Remove unused imports/variables
# Disabled: fully redundant with ruff (F401, F841, F811) + --fix
# ===========================================================================
# - repo: https://github.com/PyCQA/autoflake
# rev: v2.3.1
# hooks:
# - id: autoflake
# args:
# - --in-place
# - --remove-all-unused-imports
# - --remove-unused-variables
# - --remove-duplicate-keys
# - --expand-star-imports
# ===========================================================================
# SAFETY - Check for security vulnerabilities in dependencies
# ===========================================================================
# Note: Safety requires API key for full functionality, disabled by default
# - repo: https://github.com/Lucas-C/pre-commit-hooks-safety
# rev: v1.3.2
# hooks:
# - id: python-safety-dependencies-check
# files: requirements.*\.txt$
# ===========================================================================
# PYRIGHT - Microsoft's type checker (very strict, optional)
# ===========================================================================
# Uncomment to enable - can be slow and very strict
# - repo: https://github.com/RobertCraiworthy/pyright-action
# rev: v1.1.350
# hooks:
# - id: pyright
# ===========================================================================
# CHECK JSON/YAML/TOML formatting (runs on push only — slow Node.js startup)
# ===========================================================================
- repo: local
hooks:
- id: prettier
name: prettier (capped scope)
# Wrapper runs prettier inside its own systemd-run scope so its
# memory budget is independent of the outer pre-push cgroup that
# has already accumulated page cache from pytest/mypy/pylint.
entry: meta/scripts/run_prettier_capped.sh
language: system
types_or: [yaml, json, markdown]
exclude: >-
(?x)^(
Bash/.*|
\.venv/.*|
third_party/.*|
\.github/skills/.*|
docs/superpowers/(plans|specs)/.*|
linux_configuration/report/.*|
.*\.lock
)$
stages: [pre-push]
# ===========================================================================
# SHELLCHECK - Shell script linting
# Wrapper batches files to avoid OOM on large repos.
# ===========================================================================
- repo: local
hooks:
- id: shellcheck
name: shellcheck
# NO --severity filter: every finding shellcheck reports, down to info
# and style, fails the commit. The previous --severity=warning quietly
# accepted 222 findings across 94 of 244 shell files -- and among them
# were real defects, not noise: two scripts that reported SUCCESS on an
# unknown command, five that sourced a path off by one directory (so
# they failed at runtime), a stale `source=` directive that had silently
# disabled following for six i3blocks scripts, and a diagnostic that
# would hang on stdin when its file list came back empty.
# Settings live in .shellcheckrc (external-sources, source-path).
entry: bash -c 'printf "%s\0" "$@" | xargs -0 -n 40 shellcheck' --
language: system
types: [shell]
# shellcheck only understands sh/bash/dash/ksh; it cannot parse zsh
# syntax (e.g. extended globs like (#s)). zsh files are checked by the
# zsh-syntax hook below instead.
exclude_types: [zsh]
# Vendored upstream code: we do not own it and patches here would be
# lost on the next vendor sync. Matches the prettier hook's exclusion.
#
# PKGBUILDs are makepkg manifests, not standalone scripts: they have no
# shebang by design, and pkgname/pkgver/source/sha512sums are read by
# makepkg while pkgdir/srcdir are injected by it. shellcheck therefore
# reports every one of them as unused-or-undefined (SC2148/SC2034/
# SC2154) -- all false positives, and none suppressible without
# littering the file with directives. Same reasoning as zsh above:
# the dialect is outside what shellcheck can judge.
exclude: ^third_party/|(^|/)PKGBUILD$
# ===========================================================================
# ZSH SYNTAX - zsh -n parse check (shellcheck has no zsh mode)
# ===========================================================================
- repo: local
hooks:
- id: zsh-syntax
name: zsh syntax check
entry: bash -c 'for f in "$@"; do zsh -n "$f" || exit 1; done' --
language: system
types: [zsh]
# ===========================================================================
# CHECK PYTHON LOCATION - All Python files must be under python_pkg/
# ===========================================================================
- repo: local
hooks:
- id: check-python-location
name: check Python files are under python_pkg/
entry: meta/scripts/check_python_location.sh
language: script
types: [python]
# ===========================================================================
# REMOVE EMPTY DIRECTORIES - Clean up empty folders in the repo
# ===========================================================================
- repo: local
hooks:
- id: remove-empty-dirs
name: remove empty directories
entry: find . -type d -empty -not -path './.git/*' -delete -print
language: system
pass_filenames: false
always_run: true
# ===========================================================================
# SECRET PATTERNS - Block commits containing sensitive data
# ===========================================================================
- repo: local
hooks:
- id: check-no-secrets
name: check for leaked secrets
entry: meta/scripts/check_no_secrets.sh
language: script
exclude: ^(\.secret-patterns|\.pre-commit-config\.yaml|.*\.geojson)$
- id: check-no-dufs-conf
name: block dufs secret config (.dufs_cloud.conf)
entry: meta/scripts/check_no_dufs_conf.sh
language: script
always_run: true
pass_filenames: false
# ===========================================================================
# COMMITIZEN - Conventional commits (optional)
# ===========================================================================
# - repo: https://github.com/commitizen-tools/commitizen
# rev: v3.13.0
# hooks:
# - id: commitizen
# - id: commitizen-branch
# stages: [push]
# Four markdown namespaces (README / CLAUDE* / DOCS* / TODO*), plus the
# rule that a TODO carries "REMOVE ME AFTER FINISH" and nothing else may.
# Shared gate in ~/src/utils owns the namespaces and the exemption list, so
# one repo's idea of a valid name cannot drift from every other repo's.
- id: md-naming
name: markdown naming convention
entry: scripts/check_md_naming.sh
language: system
files: \.(md|markdown)$
# Every dependency on its ecosystem's newest STABLE release, exact-pinned
# so the question is mechanically answerable at all.
#
# `files:` is what keeps this cheap: a commit touching only source makes
# ZERO network requests, and manifest commits mostly hit a 6h disk cache.
#
# Runs WITHOUT --strict on purpose. Offline it degrades to the cached
# answer and passes with a loud banner; `--no-verify` is banned here, so
# a hook that hard-failed on a flaky connection would leave no way to
# commit at all. CI, always online, is the strict authority.
- id: dependency-freshness
name: dependencies on newest stable
entry: scripts/check_dependency_freshness.sh
language: system
pass_filenames: true
files: >-
(pubspec\.(yaml|lock)|pyproject\.toml|requirements.*\.txt|uv\.lock
|package\.json|package-lock\.json|pnpm-lock\.yaml
|Cargo\.(toml|lock)|go\.(mod|sum)
|\.fvmrc|\.nvmrc|\.python-version
|\.github/workflows/.*\.ya?ml)$