
Security News
Ruby's Bundler 4.0.18 Extends Cooldown to bundle lock and bundle cache
The supply chain control that delays freshly published gems now covers lockfile generation and gem vendoring in Ruby projects.
@touchskyer/logex
Advanced tools
Write blog-style session papers from Claude Code transcripts. One session → N articles, one per topic. The LLM (your agent) decides topic segmentation and writes the articles in-session — no extra API key needed.
logex turns Claude Code session JSONL transcripts into blog-quality technical articles. It does the boring work (parse, chunk, score, group by signal) as a pure pipeline, then hands the actual topic segmentation and writing to the LLM that's already in the loop. One long session typically yields 1–3 articles — the agent reads chunk summaries, decides what's worth writing, and produces structured JSON articles that get published to a separate data repo (logex-data).
The webapp (React + Vite SPA) reads that data repo and renders the blog at logex-io.vercel.app.
npm install -g @touchskyer/logex
Requires Node.js 18+.
/logex workflowIn Claude Code, just run:
/logex # write articles from the current session
/logex <path.jsonl> # write from a specific JSONL file
/logex --list # list recent session JSONLs
Under the hood the skill runs:
logex write # parse → chunk → score → segment (LLM) → write (LLM) → publish
logex list # show recent sessions
logex mcp # start MCP server for other agents
The agent is the one deciding topic boundaries and drafting the prose — logex is the scaffolding around it.
Install the plugin once, get the skill and the Stop-hook reminder automatically:
/plugin marketplace add iamtouchskyer/logex
/plugin install logex@logex
This registers:
logex — the /logex slash command and the full segmentation + writing procedure..claude-plugin/plugin.json — declares the skill and hook paths so the plugin travels as one unit.No global npm install required when you use the plugin — the skill ships with it.
Any MCP-capable client can drive logex via:
logex mcp
Client-side config:
| Client | Setup |
|---|---|
| Cursor | npm install -g @touchskyer/logex, then add MCP server: command logex, args ["mcp"] |
| Codex | npm install -g @touchskyer/logex && codex mcp add logex -- logex mcp |
| Windsurf / others | Same pattern — command logex, args ["mcp"] |
Articles are stored in the iamtouchskyer/logex-data repo on GitHub — logex publishes directly via the Contents API. Ensure GITHUB_TOKEN is set in ~/.claude/.env with repo scope.
MIT.
logex 把 Claude Code session 的 JSONL transcript 变成 blog 级的技术文章。它把"枯燥的部分"(parse、chunk、score、按信号分组)做成纯 pipeline,把话题切分和写作交给当前在 session 里的 LLM —— 不需要额外 API key。一次长 session 通常产 1–3 篇文章:agent 读 chunk summaries,自己决定哪些值得写,产出结构化的 JSON 文章,发布到独立的 data 仓库 logex-data。
Webapp 是 React + Vite SPA,从 data 仓库读文章,部署在 logex-io.vercel.app。
npm install -g @touchskyer/logex
需要 Node.js 18+。
/logex workflow在 Claude Code 里:
/logex # 从当前 session 写文章
/logex <path.jsonl> # 从指定 JSONL 写
/logex --list # 列最近的 session
背后调用:
logex write # parse → chunk → score → 话题切分(LLM)→ 写作(LLM)→ publish
logex list # 列最近 session
logex mcp # 给其它 agent 暴露 MCP server
话题切分和正文都是 agent 自己决定,logex 只做周边脚手架。
一次安装,skill 和 Stop hook 自动就绪:
/plugin marketplace add iamtouchskyer/logex
/plugin install logex@logex
会注册:
logex —— /logex slash command 和完整切分 + 写作流程.claude-plugin/plugin.json —— plugin 的入口,skill 和 hook 作为一个整体打包走 plugin 就不需要再 npm install -g —— skill 随 plugin 走。
任意支持 MCP 的 client:
logex mcp
客户端配置:
| 客户端 | 配置 |
|---|---|
| Cursor | npm install -g @touchskyer/logex,添加 MCP server:command logex,args ["mcp"] |
| Codex | npm install -g @touchskyer/logex && codex mcp add logex -- logex mcp |
| Windsurf / 其它 | 同样的 pattern:command logex,args ["mcp"] |
文章保存在 iamtouchskyer/logex-data 仓库(直接通过 GitHub Contents API 发布;在 ~/.claude/.env 中配好带 repo scope 的 GITHUB_TOKEN)。
MIT.
logex は Claude Code セッションの JSONL トランスクリプトを、ブログ品質の技術記事に変換します。退屈な処理(parse・chunk・score・シグナル単位でのグルーピング)はピュアな pipeline で処理し、トピック分割と執筆は、セッション内にすでにいる LLM に任せます — 追加の API key は不要。長めのセッション 1 回で通常 1〜3 本の記事が出ます。agent が chunk summaries を読み、書く価値のある話題を判断し、構造化された JSON 記事を別 repo (logex-data) に publish します。
Webapp は React + Vite SPA で、その data repo を読んで logex-io.vercel.app に表示されます。
npm install -g @touchskyer/logex
Node.js 18+ が必要。
/logex ワークフローClaude Code 内で:
/logex # 現在のセッションから記事を書く
/logex <path.jsonl> # 指定 JSONL から書く
/logex --list # 最近のセッション一覧
内部的に呼ぶコマンド:
logex write # parse → chunk → score → 分割(LLM)→ 執筆(LLM)→ publish
logex list # セッション一覧
logex mcp # 他 agent 向け MCP server
plugin を入れれば skill と Stop hook が自動で有効になります:
/plugin marketplace add iamtouchskyer/logex
/plugin install logex@logex
含まれるもの:
logex — /logex スラッシュコマンド.claude-plugin/plugin.json — skill/hook をまとめるエントリlogex mcp
| Client | 設定 |
|---|---|
| Cursor | npm install -g @touchskyer/logex、MCP server: command logex, args ["mcp"] |
| Codex | npm install -g @touchskyer/logex && codex mcp add logex -- logex mcp |
| Windsurf / その他 | command logex, args ["mcp"] |
MIT.
logex 는 Claude Code 세션의 JSONL 트랜스크립트를 블로그 품질의 기술 기사로 바꿉니다. 지루한 작업(parse, chunk, score, 시그널별 그룹핑)은 순수 pipeline 으로 처리하고, 토픽 분할과 작성은 이미 세션 안에 있는 LLM 에게 맡깁니다 — 별도 API key 불필요. 긴 세션 하나로 보통 1〜3 개의 기사가 나옵니다. agent 가 chunk summary 를 읽고 쓸 가치가 있는 것을 고른 뒤, 구조화된 JSON 기사를 별도 data repo (logex-data) 로 publish 합니다.
Webapp 은 React + Vite SPA 로, data repo 를 읽어 logex-io.vercel.app 에서 렌더링됩니다.
npm install -g @touchskyer/logex
Node.js 18+ 필요.
/logex workflowClaude Code 에서:
/logex # 현재 세션으로 기사 작성
/logex <path.jsonl> # 지정 JSONL 로 작성
/logex --list # 최근 세션 목록
내부 실행 명령:
logex write # parse → chunk → score → 분할(LLM) → 작성(LLM) → publish
logex list # 세션 목록
logex mcp # 다른 agent 를 위한 MCP server
/plugin marketplace add iamtouchskyer/logex
/plugin install logex@logex
포함:
logex — /logex slash command.claude-plugin/plugin.json — 통합 entry pointlogex mcp
| Client | 설정 |
|---|---|
| Cursor | npm install -g @touchskyer/logex, MCP: command logex, args ["mcp"] |
| Codex | npm install -g @touchskyer/logex && codex mcp add logex -- logex mcp |
| Windsurf / 기타 | command logex, args ["mcp"] |
MIT.
logex convierte los transcripts JSONL de sesiones de Claude Code en artículos técnicos con calidad de blog. La parte aburrida (parse, chunk, score, agrupación por señal) vive en un pipeline puro; la segmentación por tema y la redacción las hace el LLM que ya está en la sesión — sin API key adicional. Una sesión larga típicamente produce 1–3 artículos: el agent lee los resúmenes de chunk, decide qué vale la pena escribir y publica artículos JSON estructurados en un repo de datos aparte (logex-data).
La webapp es una SPA React + Vite que lee ese repo y se despliega en logex-io.vercel.app.
npm install -g @touchskyer/logex
Requiere Node.js 18+.
/logexEn Claude Code:
/logex # escribir artículos desde la sesión actual
/logex <path.jsonl> # desde un JSONL específico
/logex --list # listar sesiones recientes
Comandos subyacentes:
logex write # parse → chunk → score → segmentar (LLM) → escribir (LLM) → publish
logex list # listar sesiones
logex mcp # servidor MCP para otros agentes
/plugin marketplace add iamtouchskyer/logex
/plugin install logex@logex
Incluye:
logex — comando /logex y procedimiento completo.claude-plugin/plugin.json — entry point del pluginlogex mcp
| Cliente | Configuración |
|---|---|
| Cursor | npm install -g @touchskyer/logex, MCP: command logex, args ["mcp"] |
| Codex | npm install -g @touchskyer/logex && codex mcp add logex -- logex mcp |
| Windsurf / otros | command logex, args ["mcp"] |
MIT.
FAQs
Write blog-style session papers from a Claude Code session JSONL
The npm package @touchskyer/logex receives a total of 0 weekly downloads. As such, @touchskyer/logex popularity was classified as not popular.
We found that @touchskyer/logex demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago. It has 1 open source maintainer collaborating on the project.
Did you know?

Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.

Security News
The supply chain control that delays freshly published gems now covers lockfile generation and gem vendoring in Ruby projects.

Security News
During a UK cyber test, a Mythos 5 agent used sockpuppets, social engineering, and prompt injection to try to get a maintainer to merge malware.

Company News
Socket is now in the AWS Security Hub Extended plan. Adopt it through AWS, apply committed spend, and block malicious open source packages.