d80df3ed7c
范围（不含 design-story-foundation、docs/design、docs/write-chapter、
craft/、humanization/README.md 等进行中改动）:
1. humanization 规则/样例运行时数据库权威
- db/ddl/111：example_ai_flavor_rule / example_ai_flavor_sample /
example_ai_flavor_rule_event（append-only 生命周期留痕），已应用到 muse-example
- deai/load_db.py：数据库装载器，激活门/重复检测/指纹与文件装载器同源；
数据库失败关闭，不静默回退 Git 文件资产
- humanization/tools/seed_rules_db.py：YAML 种子单事务同步，幂等、
变化留痕、--strict 对 db-only 行失败关闭；真实库已种入 26 规则/107 样例
- prevent/diagnose/revise 生产路径切到数据库读取（--offline 显式读文件），
落库前新鲜度检查与合同声明来源一致；四个 SKILL.md 数据库合同同步
- 真实库验证：规则库指纹与文件种子一致（v-609bc40e21d0b5db），
三个生产脚本端到端从库装载通过
2. PostgreSQL 集成：显式授权后 9/9 通过
- 此前被依赖门阻断的 6 个 _db/smoke 测试全部通过
- extract rollback 冒烟改为回滚事务内自给夹具（pending 窗/草稿缺失时自建），
不再依赖瞬时生产状态；夹具残留核验为 0
3. Skill 行为评测脚手架（真实执行数量仍为 0）
- harness/evals/skill_eval.py：场景合同、六类评测范畴、适配器和结构化裁决报告
- diagnose-ai-flavor 参考场景 4 条 + 管道自测 7 项通过
- 真实模型适配器未授权时以稳定码 EVAL_ADAPTER_UNAVAILABLE 失败关闭；
清单登记 skill_behavior_eval 条目，默认被依赖门阻断
4. evaluate-frozen-replay raw 存储边界冲突
- docs/2026-08-19 备忘录：平台 DB-first 合同（创始人批准）与回放链
仓外 vault 强制的冲突事实、两个选项和裁决前约束；运行时合同未单方面改写
5. harness 自身修复
- runner 对账语义：行为评测入口不参与测试资产双向等值，但登记文件必须存在；
manifest 保留 skill_behavior_eval 布尔字段并校验类型
- 新增 2 条对账回归用例
验证证据: harness 三组自测 15+15+7 通过；静态审计 32 Skill / 0 问题；
76 个非数据库条目通过；9 个 PostgreSQL 集成条目显式授权后通过；
行为评测条目默认阻断；py_compile 与 git diff --check 通过。
未调用真实模型、embedding 或额度；真实行为评测执行数量仍为 0。
98 lines
4.8 KiB
Python
98 lines
4.8 KiB
Python
#!/usr/bin/env python3
|
||
"""技能 2 前置预防离线测试：负约束只取 active 规则、声音账投影、串作品拒绝。"""
|
||
import pathlib
|
||
import sys
|
||
import unittest
|
||
from unittest.mock import patch
|
||
|
||
PROJECT_ROOT = pathlib.Path(__file__).resolve().parents[3]
|
||
SCRIPT_DIR = PROJECT_ROOT / ".claude" / "skills" / "prevent-ai-flavor" / "scripts"
|
||
sys.path.insert(0, str(SCRIPT_DIR))
|
||
|
||
import prevent_ai_flavor as prev # noqa: E402
|
||
|
||
LEDGER = {
|
||
"schema_version": "voice-baseline-v1",
|
||
"work_ref": "synthetic:demo",
|
||
"narrator": {"sentence_habits": ["短句收束"], "punctuation_habits": ["少用感叹号"]},
|
||
"characters": {"老周": {"verbal_tics": ["我说小子"], "sample_lines": ["我说小子，茶要凉了。"]}},
|
||
"untouchable_verbal_tics": {"老周": ["我说小子"]},
|
||
"protected_spans": ["青云城的雨说来就来"],
|
||
"blacklist": ["值得注意的是"],
|
||
}
|
||
|
||
|
||
class PreventionContractTest(unittest.TestCase):
|
||
def test_contract_without_ledger_still_valid(self):
|
||
contract = prev.build_prevention_contract("synthetic:demo")
|
||
self.assertEqual(contract["schema_version"], "ai-flavor-prevention-v2")
|
||
self.assertGreaterEqual(len(contract["negative_constraints"]), 10)
|
||
self.assertEqual(contract["positive_samples"], [])
|
||
self.assertIsNone(contract["built_from"]["voice_ledger_sha256"])
|
||
for item in contract["negative_constraints"]:
|
||
self.assertTrue(item["avoid_examples"], item["rule_id"])
|
||
self.assertIn("keep_examples", item)
|
||
self.assertIn("boundary_examples", item)
|
||
self.assertIn("regression_traps", item)
|
||
|
||
def test_ledger_projects_positive_samples_and_blacklist(self):
|
||
contract = prev.build_prevention_contract("synthetic:demo", voice_ledger=LEDGER)
|
||
kinds = {(p["kind"], p["text"]) for p in contract["positive_samples"]}
|
||
self.assertIn(("verbal_tic", "我说小子"), kinds)
|
||
self.assertIn(("voice_sample", "我说小子，茶要凉了。"), kinds)
|
||
self.assertIn(("narrator_habit", "短句收束"), kinds)
|
||
self.assertEqual(contract["blacklist"], ["值得注意的是"])
|
||
self.assertEqual(contract["protected_spans"], ["青云城的雨说来就来"])
|
||
self.assertTrue(any("黑名单表达式" in item for item in prev.render_writer_constraints(contract)))
|
||
self.assertIsNotNone(contract["built_from"]["voice_ledger_sha256"])
|
||
|
||
def test_work_ref_mismatch_is_rejected(self):
|
||
with self.assertRaisesRegex(prev.PreventionContractError, "不一致"):
|
||
prev.build_prevention_contract("synthetic:other", voice_ledger=LEDGER)
|
||
|
||
def test_candidate_ledger_is_rejected_even_in_offline_projection(self):
|
||
bad = dict(LEDGER, status="candidate")
|
||
with self.assertRaisesRegex(prev.PreventionContractError, "canonical"):
|
||
prev.build_prevention_contract("synthetic:demo", voice_ledger=bad)
|
||
|
||
def test_bad_schema_version_is_rejected(self):
|
||
bad = dict(LEDGER, schema_version="nope")
|
||
with self.assertRaisesRegex(prev.PreventionContractError, "schema_version"):
|
||
prev.build_prevention_contract("synthetic:demo", voice_ledger=bad)
|
||
|
||
def test_cli_default_persists_run(self):
|
||
# 默认（非 --offline）模式生产读数据库规则库；离线测试补丁文件库接缝，不触真实连接。
|
||
import argparse # noqa: F401 (确认 CLI 依赖可导入)
|
||
import tempfile
|
||
import json
|
||
from deai import load
|
||
|
||
def _file_library(load_database):
|
||
samples = load.load_samples()
|
||
rules = load.load_rules(samples=samples)
|
||
return samples, rules, prev.rule_library_version(rules), "database"
|
||
|
||
with tempfile.TemporaryDirectory() as tmp:
|
||
output = pathlib.Path(tmp) / "contract.json"
|
||
with patch.object(prev, "persist_prevention", return_value={"run_id": "prev-x"}) as persist, \
|
||
patch.object(prev, "_load_db_ledger", return_value=None), \
|
||
patch.object(prev, "load_runtime_library", side_effect=_file_library):
|
||
code = prev.main(["--work-ref", "synthetic:demo", "--output", str(output)])
|
||
self.assertEqual(code, 0)
|
||
persist.assert_called_once()
|
||
contract = json.loads(output.read_text(encoding="utf-8"))
|
||
self.assertEqual(contract["work_ref"], "synthetic:demo")
|
||
|
||
def test_cli_offline_skips_db(self):
|
||
import tempfile
|
||
with tempfile.TemporaryDirectory() as tmp:
|
||
output = pathlib.Path(tmp) / "contract.json"
|
||
with patch.object(prev, "persist_prevention") as persist:
|
||
code = prev.main(["--work-ref", "synthetic:demo", "--output", str(output), "--offline"])
|
||
self.assertEqual(code, 0)
|
||
persist.assert_not_called()
|
||
|
||
|
||
if __name__ == "__main__":
|
||
unittest.main()
|