session-distill
Skill0 starsChanged 2 months agoMITView on GitHub
历史会话档案批量蒸馏 + 选稿复用——从 Claude Code / WorkBuddy / 豆包等的会话转录(jsonl)中寻找既有成稿/草稿,挖掘选题、素材卡、方法论与真实踩坑细节;并从最新素材库与舆情日报中挑出最值得发布的内容,让内容创作直接复用蒸馏产物而不用重复粘贴历史会话。当用户说分析历史会话、蒸馏会话、从会话找选题、会话挖素材、历史会话里有什么可写的、不要重复输入会话、把历史会话沉淀成技能、从历史会话找既有成稿/草稿、从素材库/舆情挑最值得发的内容时使用。触发词:会话蒸馏、蒸馏会话、历史会话分析、从会话找选题、会话挖素材、分析历史会话、会话档案挖掘、历史会话沉淀、会话找成稿、素材库选稿、舆情选稿。
1 other project uses a version of this.
Adopted by
Most recently taken up by
- youcisla/Agent-Foundrytheir own version, first seen Nov 16, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- youcisla/Agent-Foundry★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository139 other files it gives its agents.
- AGENTS.mdAGENTS.md
- skills/caveman-learn/CLAUDE.mdCLAUDE.md
- adversarial-reviewSkill
- ai-repo-engineering-layoutSkill
- ai-session-material-minerSkill
- ai-usage-reportSkill
- ardot-effect-fallbackSkill
- audio-friendly-adaptSkill
- automation-resurrectionSkill
- book2vido-videoSkill
- boss-autopilotSkill
- brand-icon-sourcingSkill
- canvas-designSkill
- cavecrewSkill
- caveman-commitSkill
- caveman-compressSkill
- caveman-discoverSkill
- caveman-evidence-reviewSkill
- caveman-exploreSkill
- caveman-helpSkill
- caveman-learnSkill
- caveman-manageSkill
- caveman-optimizeSkill
- caveman-reviewSkill
- caveman-setupSkill
- cavemanSkill
- caveman-statsSkill
- cc-switch-usage-auditSkill
- certification-submission-kitSkill
- comment-reply-playbookSkill
- concept-anchor-figureSkill
- conference-archiveSkill
- constitution-amendmentSkill
- content-format-auditSkill
- content-matrix-channel-auditSkill
- cross-repo-skill-curationSkill
- daily-judgment-briefSkill
- daily-reportSkill
- de-ai-flavorSkill
- desktop-ai-forensicsSkill
- discernment-nudgeSkill
- doc-translateSkill
- doubao-app-layer-archiveSkill
- effort-accountingSkill
- ego-browserSkill
- egress-allowlistSkill
- eval-skillsSkill
- evolve-skillsSkill
- exam-prep-pmpSkill
- fact-check-gateSkill
- fanout-mainline-ledgerSkill
- free-token-channel-auditSkill
- github-repo-gallerySkill
- git-identity-pinSkill
- git-push-in-sandboxSkill
- harness-governanceSkill
- harness-model-upgrade-auditSkill
- hidden-contract-auditSkill
- investigate-firstSkill
- ip-platform-fit-auditSkill
- The first 60 shown.
The file
Files
历史会话蒸馏(Session Distill)
一句话定位:把「历史会话档案 + 最新素材库 + 最新舆情日报」变成「最值得发布的内容」的选稿复用管线。 解决:①内容创作不用把整段历史会话再喂给 AI——直接引用蒸馏产物;②选稿不靠拍脑袋——从既有成稿/草稿与最新素材/舆情里挑出最值得发布的。 与
session-sediment分工:彼管单次会话结束时的即刻沉淀;此管历史会话档案的批量挖掘 + 选稿复用。
输入源
| 源 | 路径(各机器按实际替换) | 用途 |
| Claude Code 会话 | ~/.claude/projects/-Users-<user>-<项目>/*.jsonl | 捞成稿/草稿 + 挖料 |
| WorkBuddy 会话 | ~/.workbuddy/projects/Users-<user>-<项目>/*.jsonl | 多为当前主力,千级文件,先按 mtime 排序 |
| 豆包/千问记忆 | ~/.qwenworkcn/awareness/main/ | 记忆层,配合跨工具接续 |
| 素材库 | 项目内 素材库/(素材卡_.md / 选题_.md / 成稿/ 等子目录) | 路径 B 选稿:按 mtime 取最新条目 |
| 舆情日报 | 项目内舆情仓库(定时任务产出的最新一期) | 路径 B 选稿:取 Top 热点 |
| 发布排期表 | 项目内排期 SSoT | 交叉判断档期占用 |
若环境带会话检索工具(如 session-search:统一检索多仓会话并出高价值候选清单),先检索定位再蒸馏,省 token。
主力用法(多数场景)
两条路径,按问法命中——多数时候不是从零挖新料,而是捞既有产出挑最值得发布的。
路径 A:从历史会话找「既有成稿 / 草稿」(会话里已产出的,直接捞出来复用)
- 在 jsonl 里按成稿标题关键词 / 产物路径关键词 grep 定位——如「发布包」「建稿」「草稿箱」「定稿」「outputs/内容稿件」「归档」+ 主题词
- 命中后回读对应落盘文件,确认版本与状态(发布包 README / CHANGELOG / 操作单、草稿箱 media_id、门禁状态)
- 输出结论:可直接发 / 需打磨(列出缺什么)/ 已发(不重复),附落盘路径
路径 B:从最新素材库 + 舆情日报挑「最值得发布的内容」(选稿场景:今天/下一档发什么)
- 素材侧:素材库按 mtime 取最新素材卡/选题卡(增量优先),读选题池交叉去重
- 舆情侧:读最新一期舆情日报,取 Top 热点与可蹭角度(2-3 句结论化,URL 留附录,不在正文堆事件罗列)
- 档期侧:对发布排期表确认开放位与占用(不产"今天已发完还发什么"的怪文)
- 交叉判断:素材 × 舆情 × 档期 → 输出「最值得发布候选 ≤3 + 一句话数据理由」,撞题核查后交用户拍板;候选分级(S/A/B)口径参考团队既有选稿复盘产物
产出对接:候选落地后走团队发布 SOP 一条龙;是否更新排期由用户拍板。
次要用法:蒸馏挖新料(选题/素材卡/方法论/踩坑)
1. 定位会话子集
按主题关键词 / 时间范围 / 文件 mtime 过滤候选 jsonl(先 ls -t 看 mtime 或 grep -l 命中文件,不裸扫全库——千级文件全扫费 token)。
2. 抽取用户消息(过滤注入)
只取 role == user 且 type == input_text 的文本;剔除 <system-reminder、<cb_summary、<conversation_history_summary、<task-notification 注入段——这些是框架/任务注入,不是用户原话,混入会污染蒸馏结论。
3. 主题聚类与词频
对用户原话做词频统计(2-6 字中文连续串)+ 主题关键词计数,识别反复出现的需求/痛点/决策场景。词频结论须能回指具体会话(附会话 ID)。
4. 蒸馏产出(对齐内容护城河)
- 选题卡:拍前三问(对谁有帮助/什么场景能用/当不当资产留下)+ 反常识点 + 数据钩子 → 追加选题池
- 素材卡:单条高价值素材(方法论/金句/可复用思路),带出处(会话 ID + 时间戳)→ 素材库
- 踩坑细节:优先提取真实一线项目细节(时间点/排查时长/独有解法)——内容护城河(拒绝 AI 流水线二创套话),可配 1 张真实界面截图(打码)
- 方法论沉淀:反复出现的流程/规则 → 视价值沉淀为新 skill(先查用户实际用词再命名)或团队铁律
5. 落盘 + 登记
| 产物 | 落点(按各项目实际仓库结构) |
| 选题卡 | 内容仓库 01_灵感与选题池/ |
| 素材卡 | 素材库/素材卡_<主题>_<日期>.md + 素材库索引 |
| 选稿候选/复盘 | 项目 outputs/选稿综合判断_<日期>.md 或选稿复盘 |
| 会话复盘 | 内容仓库 04_会话与复盘/ |
| 留痕 | 项目 CHANGELOG + 产出索引 |
纪律
- 只读挖掘:不改会话档案、不删文件(结构性删除走项目安全删除工具,禁
rm硬删) - 诚实讲局限:抽取不到就写「无」,不编造会话内容
- 事实可溯源:每条素材附会话 ID 路径可回溯,数字标注来源+验证程度
- 脱敏:蒸馏产物若上公网/开源,先过隐私扫描(如
privacy_scan --fail-fast),真名/手机号/邮箱/雇主信息一律模糊化 - 选稿口径:候选必须给「一句话数据理由」,不发拍脑袋内容;已发内容不重复建议
- 适用范围:工作区专属技能须带范围守卫(读根
AGENTS.md含空间标识才执行),其他项目不适用
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

