liqiang
b119135836
Build latest book artifacts / build (push) Canceled after 0s
dependency resolution / resolve (3.11) (push) Canceled after 0s
dependency resolution / resolve (3.13) (push) Canceled after 0s
deploy-pages / build (push) Canceled after 0s
deploy-pages / deploy (push) Canceled after 0s
i18n consistency check / check (push) Canceled after 0s
provider adoption tests / test (chapter2/context-compression) (push) Canceled after 0s
provider adoption tests / test (chapter2/prompt-injection) (push) Canceled after 0s
provider adoption tests / test (chapter2/system-hint) (push) Canceled after 0s
provider adoption tests / test (chapter3/log-sanitization) (push) Canceled after 0s
web-search-agent tests / test (push) Canceled after 0s
web-search-agent tests / agentbook (push) Canceled after 0s
844 B
844 B
无编号补充案例 结果（2026-08-07）
执行结果
- 16 项机制测试全部通过。
- 离线路径：21 条候选规则合并为 8 条；未完成任务集检出 8/8；正常文本保留集误伤 0/8；金标校准 10/10。
- 第一次真实 LLM 运行直接采用模型规则，结果为检出 0/8、误伤 7/8，说明模型提出的规则不能未经检查直接进入 Skill。
- 修改后，模型外部代码只接纳能映射到受支持检测器的候选；无法映射时回退到确定性规则。最终真实路径为 21 条候选合并成 8 条，检出 8/8、误伤 0/8、金标校准 10/10，整体验收通过。
最终证据见 real_20260807T160547Z/evidence.json。该结果说明，LLM 适合提出写作规则候选，但规则合并、检测器选择、校准和发布仍应由模型外部代码控制。