kangshifu-board/scripts/audit-release-workspace.py

"""Inventory source changes without including private data, generated releases or retired demos."""
import hashlib
import json
from pathlib import Path
import re
import subprocess

ROOT = Path(__file__).resolve().parents[1]
ALLOWED = ('scripts/', 'docs/', 'deploy/', 'integrations/', 'server/', 'skills/', 'test/')

def main():
    pending = subprocess.check_output(['git', 'ls-files', '--others', '--exclude-standard', '-z'], cwd=ROOT).decode('utf-8').split('\0')
    for args in (['git', 'diff', '--name-only', '-z'], ['git', 'diff', '--cached', '--name-only', '-z']):
        pending += subprocess.check_output(args, cwd=ROOT).decode('utf-8').split('\0')
    candidates = sorted({p for p in pending if p and (p.startswith(ALLOWED) or p in {'.gitignore', '.gitattributes', 'package.json'})})
    findings = []
    forbidden = []
    for name in candidates:
        p = ROOT / name
        if not p.exists():
            continue
        if p.is_symlink() or p.suffix.lower() in {'.zip', '.sqlite', '.db', '.pem', '.key', '.pyc', '.png'} or name.endswith('credential-binding.json'):
            forbidden.append(name)
            continue
        text = p.read_text(encoding='utf-8-sig')
        for m in re.finditer(r'''(?i)\b(master_?key|service_?token|access_?token|api_?key|client_?secret)\b\s*[:=]\s*['"]([^'"\n]{16,})['"]''', text):
            if not ('test/' in name or 'server/test/' in name or Path(name).name.startswith('test_')):
                findings.append({'path': name, 'line': text.count('\n', 0, m.start()) + 1, 'field': m[1], 'literalLength': len(m[2])})
        if re.search(r'-----BEGIN (?:RSA |OPENSSH )?PRIVATE KEY-----\s+[A-Za-z0-9+/]{40}', text):
            forbidden.append(name)
    skills = {}
    for skill in (ROOT / 'skills').iterdir():
        manifest = skill / 'FILES.sha256'
        if not manifest.is_file():
            continue
        rows = manifest.read_text(encoding='utf-8').splitlines()
        for line in rows:
            expected, relative = line.split('  ', 1)
            actual = hashlib.sha256((skill / relative).read_bytes()).hexdigest()
            if expected != actual:
                raise ValueError('SKILL_MANIFEST_MISMATCH:' + skill.name + '/' + relative)
        skills[skill.name] = {'filesVerified': len(rows), 'manifestSha256': hashlib.sha256(manifest.read_bytes()).hexdigest()}
    out = {'ok': not forbidden and not findings, 'sourceCandidates': candidates, 'sourceCandidateCount': len(candidates),
           'forbiddenCandidates': forbidden, 'literalCredentialCandidates': findings, 'skills': skills,
           'excludedLocalOutputs': ['artifacts/', 'data/', 'work/', 'private database files', 'credentials', 'retired demo drafts'],
           'productionWrites': 0}
    target = ROOT / 'work/complete-release-20261106'
    target.mkdir(parents=True, exist_ok=True)
    (target / 'workspace-audit.json').write_text(json.dumps(out, ensure_ascii=False, indent=2) + '\n', encoding='utf-8')
    print(json.dumps({k:v for k,v in out.items() if k != 'sourceCandidates'}, ensure_ascii=False, indent=2))

if __name__ == '__main__':
    main()