dqx-conventional-commit
by nampham2
Data quality as code. Works with your warehouse, scales with your needs.
Run in ManusSKILL.md
name: dqx-conventional-commit description: Create conventional commit following DQX standards compatibility: opencode metadata: workflow: version-control audience: all-agents
What I do
Guide you through creating a properly formatted conventional commit that passes DQX's commitizen validation.
Commit Format
<type>(<scope>): <subject>
<body>
<footer>
Required Parts
Type (required): Category of change Scope (optional): Area of codebase affected Subject (required): Brief description (max 72 characters)
Optional Parts
Body: Detailed explanation (why, not what) Footer: Issue references, breaking changes
Commit Types
Choose the type that best describes your change:
| Type | When to Use | Example |
|---|---|---|
| feat | New feature | feat(cache): add LRU cache with TTL support |
| fix | Bug fix | fix(evaluator): resolve type inference issue |
| docs | Documentation only | docs(api): update MetricProvider examples |
| test | Adding/updating tests | test(parser): add coverage for edge cases |
| refactor | Code changes (no bugs/features) | refactor(graph): simplify traversal algorithm |
| perf | Performance improvements | perf(analyzer): optimize SQL generation |
| style | Formatting changes only | style(api): fix line length violations |
| build | Build system/dependencies | build(deps): upgrade pyarrow to 15.0.0 |
| ci | CI configuration | ci(GitHub): add coverage reporting |
| chore | Maintenance tasks | chore(cleanup): remove deprecated code |
| revert | Revert previous commit | revert: feat(cache): add LRU cache |
DQX-Specific Scopes
Use these scopes for DQX modules:
Core modules:
analyzer- SQL analysis engineapi- User-facing APIgraph- Graph processingevaluator- Expression evaluationprovider- Metric providerspecs- Assertion specificationsvalidator- Validation logic
Supporting modules:
display- Terminal UI with Richorm- Object-relational mappingextensions- Extensions systemcommon- Shared utilitiesdialect- SQL dialects (DuckDB, BigQuery, Snowflake)functions- Built-in functionsmodels- Data modelsops- Operationsstates- State managementutils- Utility functionsdql- Data Quality Language parser
Infrastructure:
deps- Dependenciesconfig- Configurationdocs- Documentationtests- Testing infrastructureci- CI/CDagents- AI agent configurationskills- OpenCode skills
Subject Line Rules
✅ Do:
- Use imperative mood ("add" not "added")
- Start with lowercase
- No period at the end
- Max 72 characters
- Be specific and descriptive
❌ Don't:
- Use past tense ("added", "fixed")
- Start with capital letter
- End with period
- Be vague ("update code", "fix bug")
Examples
✅ Good:
feat(cache): add LRU cache with TTL support
fix(parser): handle empty input strings
docs(api): clarify MetricProvider usage
test(graph): add traversal edge cases
❌ Bad:
feat(cache): Added cache. # Past tense, capital, period
fix: fixed a bug # No scope, vague, past tense
Update documentation # No type, capital, vague
feat(cache): Implement new caching system for metrics. # Capital, period, too long
Body Guidelines
When to add body:
- Complex changes need explanation
- Non-obvious reasons for change
- Breaking changes
- Multiple related changes
Format:
- Separate from subject with blank line
- Wrap at 72 characters
- Explain why, not what
- Use bullet points for multiple items
Example:
feat(analyzer): add query optimization for large datasets
Improve performance for queries with >1M rows by:
- Implementing query result caching
- Adding early termination for LIMIT queries
- Optimizing JOIN operations
Benchmarks show 3x speedup on typical workloads.
Footer Guidelines
Issue references:
Closes #123
Fixes #456
Resolves #789
Refs #101
Breaking changes:
BREAKING CHANGE: remove deprecated `check_all()` method
Use `run_all()` instead. Migration:
- Old: suite.check_all()
- New: suite.run_all()
Complete Examples
Simple feature
feat(cache): add LRU cache with TTL support
Bug fix with context
fix(evaluator): resolve type inference for nested expressions
The type inference was failing for expressions with more than
two levels of nesting due to incorrect visitor traversal order.
Fixes #234
Breaking change
feat(api): redesign MetricProvider constructor
Simplify API by accepting config dict instead of individual params.
BREAKING CHANGE: MetricProvider constructor signature changed
Old:
provider = MetricProvider(host, port, db, user, password)
New:
provider = MetricProvider(config={
"host": host,
"port": port,
"database": db,
"credentials": {"user": user, "password": password}
})
Migration guide: docs/migration-v2.md
Closes #456
Multiple related changes
refactor(graph): improve visitor pattern implementation
- Extract base visitor class for reusability
- Add type hints for all visitor methods
- Simplify traversal logic with generator pattern
- Add comprehensive docstrings
This refactoring improves code maintainability and makes it
easier to implement custom visitors for new use cases.
Validation
Your commit message is validated by commitizen hook during pre-commit:
uv run pre-commit run --all-files
Common validation errors:
❌ subject must start with lowercase:
- Fix: Change first letter to lowercase
❌ subject is too long (>72 chars):
- Fix: Shorten subject, move details to body
❌ type must be one of [feat, fix, docs, ...]:
- Fix: Use valid type from list above
❌ subject must not end with period:
- Fix: Remove trailing period
When to use me
Use this skill when:
- Committing code after each implementation phase
- Addressing feedback with fix commits
- Unsure about format for your change type
- Commit rejected by pre-commit hook
Quick Decision Tree
Adding new functionality? → feat(scope):
Fixing a bug? → fix(scope):
Updating documentation? → docs(scope):
Adding/updating tests? → test(scope):
Refactoring code? → refactor(scope):
Improving performance? → perf(scope):
Formatting only? → style(scope):
Dependencies/build? → build(deps):
Maintenance? → chore(scope):
Reference
Complete details: AGENTS.md §commit-conventions
Commitizen validation runs automatically during pre-commit hooks.
Score
Total Score
Based on repository quality metrics
SKILL.mdファイルが含まれている
ライセンスが設定されている
100文字以上の説明がある
GitHub Stars 100以上
3ヶ月以内に更新
10回以上フォークされている
オープンIssueが50未満
プログラミング言語が設定されている
1つ以上のタグが設定されている
Reviews
Reviews coming soon
