昆仑生态技能 — 分发名:
kunlun-output-packaging| 版本:v2.1.1 与kunlun-dashboard(仪表盘快照)和kunlun-knowledge-structure(insight card入库)衔接。
Output Packaging Checklist
**v2.2.0 更新(2026-05-28):同步 Q-Bridge v1.0 — 协议A→凤口路由卡、协议C→凤口输出协议、eleven-bridges→knowledge-tree v2.0 更新(2026-05-26):
- SOUL.md 共振腔架构 v3.0 → 本检查清单对齐认知知识图+三视图投影
- 更新对齐认知知识图+三视图投影
- 解冻(frozen→active),成为认知知识图三视图出口的可选补充工具
The Core Problem
After a deep analysis, there are typically 4-7 files that need synchronized updates. Manual tracking has a reliability ceiling. This checklist eliminates that.
Mandatory Checklist (Run Every Time)
Step 0: 输出准备的预处理
- [ ] 分析视图的结构已构建完成?含OCGS Core回路、域适应三棱镜、免疫系统证伪、蒸馏塔抽象、窗口赛跑记录、认知盲点
- [ ] 是否需要格式渲染(对外呈现)?
- 需要 → 产出类型自判+格式渲染+交付(默认自动执行)
- 不需要 → 直接以MD交付
Step 0: Knowledge Source Integrity Check
- [ ] 本次分析的推理依据是否全部标注了信度等级?
- 如果发现未标注 → 补标
[T1]/[T2]/[T3] - 如果发现使用了T3知识作为推理依据 → 声明"此段判断未经思想库验证"
- 如果发现未标注 → 补标
- [ ] 本次分析产出的新知识 → 标注[T3 待验证]
- 如果是验证了已有的T3知识 → 标记升T2候选
③学科响应网络激活声明(v2.1 取代原Bridge Activation Statement)
在产出封装前,记录本次分析的桥激活状态——这是凤口(桥路由)→龙门(挂桥入库)的必填环节。
- [ ] 桥激活声明已填写 → 存入分析报告 §2 桥层声明
- P0(皇帝):[桥编号] — 得分+理由
- P0(宰相):[桥编号] — 得分+理由(可选)
- P1(顾问):[桥编号] — 得分+理由(可选)
- 跨域映射触发:[M-xxx, M-yyy](凤口桥共振自动扫描)
- Gap桥呼唤:[桥编号] → 呼唤计数+[N](当前总计=[])
- 桥间冲突仲裁:[桥组合] → Layer 1/2/3判断
Cognitive Blind Spot (New Mandatory Step)
在完成分析框架和结论后、填写交付清单前,先花30秒记录本次分析的认知盲区。
- [ ] 认知盲点已填写 → 存入分析报告 §9
- 信息缺失(什么关键数据没拿到)
- 视角缺失(什么重要视角被忽略)
- 被迫假设(因证据不足而做的假设)
- 免疫系统条件标记缺失(哪些结论缺少证伪验证,原因是什么)
Primary Deliverables
- [ ] Summary — 5-line condensed version output in the chat for immediate consumption
- [ ] 发布审计卡(如果本次产出涉及版本发布)— 使用
memory/change-reports/YYYY-MM-DD_release-vX.Y.Z.md模板- 包含:变更摘要、版本变化、预检结果、安全修复清单、已知限制、回退方案
- 自动触发:
release-preflight.sh --ci预检 +release-checklist.md手工确认 - 审计卡作为发布流程的最终输出物(§9.2 [5]),写入
memory/change-reports/
- [ ] Analysis report — Archived to memory/analysis/ using TEMPLATE.md format
- Include: problem framework, OCGS structures, circuits, base patterns, BCS stage determination
- [ ] Insight cards — Written to memory-index.md with proper section tagging
- Each insight now includes 信度等级 field (T1/T2/T3/T4)
- Use 规范引用别名 in the "对应memory-index条目" column of the insight card table
- ⚡ Auto-run:
python3 memory/maintain_counters.py count— reads the latest report's insight card table, matches aliases, updates counts
Knowledge Trust & Cards
- [ ] New knowledge entry → determine trust level:
- T3 (learning): write to memory-index.md → optionally create study-card if new discipline
- T2 (verified): write to memory-index.md → create axiom-card or case-card
- T1 (axiom): write to MEMORY.md → update axiom-cards-index.md
- [ ] If new discipline involved → create study card in
memory/views/study-cards-index.md - [ ] If T2+ knowledge → update axiom-cards-index.md or case-library-index.md
- [ ] Cross-domain validation — When this is the second+ analysis in a family, compare structures
Ecosystem Synchronization
-
[ ] Eco health check — Run
python3 kunlun-dist/scripts/ecosync-check.pyto detect version drift- If '❌ 有差异': decide whether sync is needed → run
ecosync-sync.sh - If drift is intentional (e.g. deployed ahead of dist): update SKILLS-VERSION.md
- If '❌ 有差异': decide whether sync is needed → run
-
[ ] VERSION.md — Updated with new version number and change summary
-
[ ] MEMORY.md — Status section synchronized (completed/in-progress/blocked/todo)
-
[ ] Daily log — New cognitive event entry added to memory/YYYY-MM-DD.md
-
[ ] 凤口输出回喂 — 分析闭环后回喂分流:
- 跨桥交叉产物 → 桥积累库(
knowledge-tree/bridge-memory-integration.md通道D) - 可复用法则/命题 → 记忆系统(MEMORY.md + memory-index.md)
- 单系统内条目碰撞 → 共鸣网络(
resonance-net.mdRP#编号) - 缺口桥标记 → 呼唤次数+1(如有)
- 跨桥交叉产物 → 桥积累库(
-
[ ] Bridge state update — 本次激活的桥 {Qxx, Qyy} 各强度+1,最后激活时间更新
-
[ ] Anti-fragility pool — New strong counter-examples added to anti-fragility-pool.md
-
[ ] Dashboard snapshot — Run
python3 generate_dashboard.pyto regenerate dashboard.html with latest state⚠️ The VERSION.md must be updated before regenerating the dashboard — the dashboard reads the version number from VERSION.md's
当前版本:field.
Optional (by need)
- [ ] 格式渲染 → 通过内容产线路由表调度(xiaoyi-report / xiaoyi-ppt / xiaoyi-xlsx / xiaoyi-pdf),不走独立格式路由
- [ ] Skill knowledge sync (if analysis affected any kunlun-dist skill contents)
Typical Execution Order
Knowledge Source Integrity Check
学科响应网络激活声明(v2.1)— 特征向量匹配度+自激活学科列表+跨域映射
Summary (chat)
Report archive
Insight cards + trust level + anti-fragility pool
Cards (study-card / axiom-card / case-card)
凤口输出回喂 — 分流桥积累/记忆系统/共鸣网络
Bridge state update — 强度+1 / 缺口呼唤+1
Cross-domain validation
Version + MEMORY + daily log
Dashboard snapshot
References
kunlun-pivot-essentials.md— 昆仑枢要(完整指南)memory/views/axiom-cards-index.md— 公理卡视图memory/views/study-cards-index.md— 学习卡视图memory/views/case-library-index.md— 案例库视图references/TEMPLATE.md— The canonical analysis report template
Step 6: 质量回映(认知知识图三视图投影后强制)
- [ ] 执行回映三问
- [ ] Q1: 回映指数(1-5分)— 回答最初问题了吗?
- [ ] Q2: 薄弱自识别 — 最弱环节是什么?
- [ ] Q3: 工序跳过声明 — 跳过了哪步?原因?
- [ ] 自评置信度判定
- [ ] 高 → 独立交付
- [ ] 中 → 标记“建议乾坤抽检”
- [ ] 低 → 强制召唤乾坤复核
- [ ] 质量回映区块写入分析报告末尾
- [ ] 回映指数计入 learning-progress.md(月度趋势追踪)
微信扫一扫