release-prep（发布准备）

release-prep — Release Preparation & Code Audit

Deep automated analysis of code quality, documentation completeness, and release readiness for Python packages. Can auto-fix issues found.

Modes

• audit (default) — analyze and report only
• fix — analyze, then auto-fix issues via coding agent
• release — full pipeline: audit → fix → changelog → version bump → tag → publish

Parse mode from user request. Default to audit if unclear.

Phase 1: Deep Code Audit

Run all checks, collect results into a structured report.

1.1 Test Suite

# Run full test suite with coverage
python -m pytest tests/ -q --tb=short --co -q 2>/dev/null | tail -1  # count tests
python -m pytest tests/ -q --tb=short 2>&1 | tail -5                  # run tests

Report: total tests, passed, failed, skipped, coverage % (if pytest-cov available).

BLOCKER if any test fails.

1.2 Static Analysis

# Ruff (if available)
ruff check src/ --statistics 2>&1 | tail -20

# Mypy (if available)
mypy src/ --no-error-summary 2>&1 | grep "error:" | wc -l

Report: error count by category. BLOCKER if errors > 0 (warnings are OK).

1.3 Dead Code Detection

# Find unused imports
ruff check src/ --select F401 2>&1

# Find functions with no callers (heuristic)
grep -rn "^def \|^async def " src/ --include="*.py" | while read line; do
  func=$(echo "$line" | sed 's/.*def \([a-zA-Z_]*\).*/\1/')
  if [ "$func" != "__init__" ] && [ "$func" != "__" ]; then
    count=$(grep -rn "$func" src/ --include="*.py" | grep -v "^def \|^async def " | wc -l)
    if [ "$count" -lt 2 ]; then
      echo "POSSIBLY UNUSED: $line"
    fi
  fi
done

Report: list of potentially dead code. WARNING level.

1.4 API Completeness

# Public API (__all__ exports) vs actual public functions
python3 -c "
import ast, sys, pathlib
for f in pathlib.Path('src/').rglob('*.py'):
    tree = ast.parse(f.read_text())
    all_list = [n.value.s for n in ast.walk(tree) 
                if isinstance(n, ast.Assign) 
                for t in n.targets if isinstance(t, ast.Name) and t.id == '__all__'
                for elt in n.value.elts if isinstance(elt, ast.Constant)]
    if all_list:
        funcs = [n.name for n in ast.walk(tree) 
                 if isinstance(n, (ast.FunctionDef, ast.AsyncFunctionDef)) 
                 and not n.name.startswith('_')]
        missing = [f for f in funcs if f not in all_list]
        if missing:
            print(f'{f}: public but not in __all__: {missing}')
"

Report: functions missing from __all__. WARNING level.

1.5 Dependency Check

# Check for pinned vs unpinned deps
grep -E "dependencies|requires" pyproject.toml | head -20

# Check for unused dependencies (heuristic)
for dep in $(python3 -c "
import tomllib; 
d=tomllib.load(open('pyproject.toml','rb')); 
print(' '.join(d.get('project',{}).get('dependencies',[])))
"); do
  pkg=$(echo "$dep" | sed 's/[>=<].*//')
  count=$(grep -rn "import $pkg\|from $pkg" src/ --include="*.py" | wc -l)
  if [ "$count" -eq 0 ]; then
    echo "POSSIBLY UNUSED DEP: $dep"
  fi
done

Report: dependency issues. WARNING level.

Phase 2: Documentation Audit

2.1 Docstring Coverage

python3 -c "
import ast, pathlib
total = missing = 0
for f in pathlib.Path('src/').rglob('*.py'):
    tree = ast.parse(f.read_text())
    for node in ast.walk(tree):
        if isinstance(node, (ast.FunctionDef, ast.AsyncFunctionDef, ast.ClassDef)):
            if not node.name.startswith('_') or node.name == '__init__':
                total += 1
                if not ast.get_docstring(node):
                    missing += 1
                    print(f'  MISSING: {f}:{node.lineno} {node.name}')
print(f'\
Docstring coverage: {(total-missing)/total*100:.0f}% ({total-missing}/{total})')
"

WARNING if coverage < 80%. BLOCKER if < 50%.

2.2 README vs Reality

Check that README.md mentions:

• All CLI entry points from pyproject.toml [project.scripts]
• Installation instructions
• Basic usage example
• All major features (compare against module-level docstrings)

# Extract entry points
python3 -c "import tomllib; d=tomllib.load(open('pyproject.toml','rb')); [print(k) for k in d.get('project',{}).get('scripts',{}).keys()]"

# Check README mentions them
for ep in $(python3 -c "import tomllib; d=tomllib.load(open('pyproject.toml','rb')); [print(k) for k in d.get('project',{}).get('scripts',{}).keys()]"); do
  grep -q "$ep" README.md && echo "✅ $ep in README" || echo "❌ $ep NOT in README"
done

WARNING if entry points missing from README.

2.3 Changelog Check

# Check CHANGELOG.md exists and has recent entries
if [ -f CHANGELOG.md ]; then
  head -20 CHANGELOG.md
  echo "Last entry date:"
  grep -oE '[0-9]{4}-[0-9]{2}-[0-9]{2}' CHANGELOG.md | head -1
else
  echo "❌ No CHANGELOG.md"
fi

2.4 Version Consistency

# Check version in pyproject.toml matches __version__
PYPROJECT_VER=$(python3 -c "import tomllib; print(tomllib.load(open('pyproject.toml','rb'))['project']['version'])")
CODE_VER=$(grep -r "__version__" src/ --include="*.py" | head -1 | grep -oE '[0-9]+\.[0-9]+\.[0-9]+')
echo "pyproject.toml: $PYPROJECT_VER"
echo "Code __version__: $CODE_VER"
if [ "$PYPROJECT_VER" != "$CODE_VER" ]; then
  echo "❌ VERSION MISMATCH"
fi

BLOCKER if version mismatch.

Phase 3: Report

Generate structured report:

╔══════════════════════════════════════╗
║     RELEASE READINESS REPORT        ║
╠══════════════════════════════════════╣
║ Package: {name} v{version}          ║
║ Branch:  {branch}                   ║
║ Commit:  {short_sha}                ║
╠══════════════════════════════════════╣
║ 🔴 BLOCKERS: {count}               ║
║ 🟡 WARNINGS: {count}               ║
║ 🟢 PASSED:   {count}               ║
╠══════════════════════════════════════╣
║ Tests:      {pass}/{total} ✅/❌    ║
║ Lint:       {errors} errors         ║
║ Docstrings: {pct}% coverage         ║
║ README:     {status}                ║
║ Version:    {status}                ║
║ Changelog:  {status}                ║
╚══════════════════════════════════════╝

BLOCKERS:
  1. {description}
  2. {description}

WARNINGS:
  1. {description}
  2. {description}

VERDICT: 🟢 READY / 🟡 READY WITH WARNINGS / 🔴 NOT READY

If mode is audit — stop here, present report to user.

Phase 4: Auto-Fix (mode=fix or mode=release)

For each issue found, categorize fix approach:

Rules:

• Show user what will be fixed before doing it
• Delegate complex fixes (test failures, missing features) to coding agent
• Simple fixes (lint, imports, version sync) — do directly
• After fixes — re-run Phase 1+2 to verify

Phase 5: Changelog Generation (mode=release)

LAST_TAG=$(git describe --tags --abbrev=0 2>/dev/null || echo "")
if [ -n "$LAST_TAG" ]; then
  RANGE="${LAST_TAG}..HEAD"
else
  RANGE="HEAD~50..HEAD"
fi

git log $RANGE --pretty=format:"%s" | sort

Group by conventional commit type (feat/fix/refactor/docs/test/chore). Generate markdown. Prepend to CHANGELOG.md.

Show draft to user before writing.

Phase 6: Version & Publish (mode=release)

1. Determine version bump (major/minor/patch) from changelog:

- feat → minor, fix → patch, BREAKING CHANGE → major

1. Update version in pyproject.toml and __version__
2. Commit: chore: release v{version}
3. Tag: v{version}
4. Build: uv build or python -m build
5. Ask user before publish: show summary + ask confirmation
6. Publish: uv publish or twine upload dist/*
7. Push tag: git push origin v{version}

Project Detection

Auto-detect project type from files present:

• pyproject.toml → Python package (primary)
• package.json → Node.js (future)
• Cargo.toml → Rust (future)

Read pyproject.toml for: name, version, entry points, dependencies, build system.

Error Handling