一句话定位

geml 是一个 DSH 配置 bundle，把 GEML 的"按块读写文档"能力装进 harness——模型不再整篇读写 Markdown/SKILL 文件，而是按小节精确读取和改写，省下大量 Token；附带一个调用图谱技能，让 Agent 能像翻书一样浏览项目的代码结构。

核心能力

• 按块读取长文档：传入块地址，只返回那一节内容，无需把整篇文档塞进上下文
• 按块改写文档：用地址定位修改点，原文件不会被改写格式，写不匹配会直接报错
• 把任意 Markdown 当 GEML 读：无需转换，原 Markdown 文件直接当只读层用
• 构建并浏览项目调用图谱：识别 TS、Rust、Java、C、Python、Go 等语言，生成谁调用谁的影响链路，浏览器实时渲染
• 把 MCP 服务限定在会话项目目录：每个会话独立起一个 stdio 进程，访问范围不互相串
• 提供两份技能文档：写作技能教 Agent 按块工作，代码图谱技能教 Agent 看/建/更新调用图

技术实现

• 语言: TypeScript / YAML（配置 bundle，无自有运行时代码；MCP 服务与技能是 dsh/CLI 自带）
• 关键依赖: @deepseek-ai/dsh-mcp-client（MCP 启动器）、@deepseek-ai/dsh-skill-filesystem（技能加载器）、@geml/geml（通过 npx 调用的 CLI）
• 架构模式: DSH cordis 配置注入——cordis.patch.yml 通过 dsh.bundle.patch 字段挂到宿主，向 cordis 配置树插入一行 MCP 客户端和一行技能文件系统，宿主在启动时合并；技能目录用 baseUrl 相对解析，bundle 装到哪技能就跟到哪
• 入口文件: cordis.patch.yml（无运行时入口，配置即行为）

适用场景

当 Agent 在多轮交互里反复读改同一份长 Markdown/SKILL/规范文档时，整篇读写会让上下文迅速膨胀并偏离事实。本插件让模型按"块地址"精确读写，仅来回一小节内容，省 Token、保留原格式。它也适合做大型项目的代码导航：让 Agent 自动识别项目语言、生成调用图、用浏览器看模块/方法的调用链路，不再靠 grep 猜关系。

前置依赖与兼容性

安装方式

dsh plugin --profile web add github:geml-spec/geml/integrations/dsh-plugin

配置项

bundle 默认配置即可使用。如需自定义，在 profile 的 cordis.patch.yml 里按 id 重写对应行（必须把所有键重新写全）：

常见问题

Q: 这个插件是干什么用的？和 VS Code 的扩展是一回事吗？

A: 不完全是代码层面的"扩展"。它向 DSH 注入一份配置，让 Agent 能用 GEML 协议按"块"读写长文档，并附带一份代码调用图谱技能。你可以理解为给 Agent 装上"按章节翻文档 + 看代码谁调用谁"两件工具。

Q: 我不想改 Markdown 格式，能直接用吗？

A: 能。GEML 把普通 Markdown 当只读层使用——geml list/find/get 直接读 .md，原文件保持原样不转换；只有当你主动把项目迁到 GEML（新建 .geml 文档）时才会出现 .geml 格式。

Q: 怎么验证插件装好了？

A: 跑 dsh --profile web --dump-config，应该能看到一段 # == @geml/dsh-plugin 的配置层；再启动 dsh --profile web，模型会话里会出现 mcp__geml__* 系列工具和 geml / geml-code-graph 两份技能。

Q: 我能把 MCP 钉到某个具体版本，避免 npx 拉最新版出问题吗？

A: 可以。在你 profile 的 cordis.patch.yml 里覆盖 id: mcp-geml 那一行，把 args 里的 @geml/geml 改成 @geml/[email protected] 这种带版本的形式即可；也可以用 PATH 上的全局 geml 命令，把 command 从 npx 改成 geml 并去掉 npx 相关参数。

Q: 调用图谱技能支持哪些语言？

A: TypeScript/JavaScript（用 scip-typescript）、Rust（rust-analyzer scip）、Java 与 C（用 Joern，需要 JDK）、Python/Go/Kotlin（Joern，可用但精度次之）；Vue/Svelte 单文件组件走自动构建路径。检测由 Agent 自动完成，不会问用户。

Q: 卸载干净吗？会留下什么残留配置吗？

A: dsh plugin --profile web remove @geml/dsh-plugin 会同时移除依赖项和注入的 cordis 这一层，profile 回到安装前的状态，无残留。

Q: 报错说找不到 geml 命令怎么办？

A: bundle 通过 npx 拉取 @geml/geml CLI 启动 MCP；如果你禁用了 npx 或网络受限，应先在 PATH 装好 geml，然后在 cordis.patch.yml 里把 command 改为 geml 并去掉 -y @geml/geml mcp 那几个参数。

上手难度

入门 — 装一行命令即生效；用户无需理解 cordis 配置语法，默认配置就能让模型使用所有能力；如需钉版本或换全局 CLI 才需要改 cordis.patch.yml。

已知问题与限制

• 非 GEML 文件上的 geml find 需要 @geml/geml 1.7.5 及以上版本；旧版本对非 .geml 文件的搜索会静默返回空结果而非报错，遇到"该命中的搜索没结果"时改用 geml list
• 调用图谱对 Java/C 依赖外部工具 Joern；Joern 需要 JDK 且首次使用时要按提示定位安装目录并写入 ~/.claude/skills/geml-code-graph/config.json
• 调用图谱对前端框架存在已知盲区：Vue/React 中组件 TAG 用法（如 <Child/>）不会被识别为调用边；Nuxt 自动导入（未显式 import 的 ref、自动注册组件）也无法识别
• React 间接派发（回调 prop 调用、dispatch()→reducer 处理、context 注入函数）以及 memo()/forwardRef() 包装的组件不会出现在调用边里，需要靠 grep 兜底
• scip-typescript 对顶级 <script setup> 调用（含 computed(() => …) 体内）会丢失，类似普通 TS 的模块级调用
• 售卖/嵌入的源码树（如 next.js 的 packages/next/src/compiled/）会大幅膨胀构建任务清单，需用 --exclude "src/compiled/**" 排除

GEML — General Expressive Markup Language

English | 中文

GEML is an Agent-Native fundamental markup format and protocol, designed for people and AI agents to read and write the same document.
One format, two readers. In agent-driven development and knowledge work, plain text and Markdown have no deterministic block boundaries: a program and a model trade the whole file in and the whole file back out — at best probing for it with line windows, and restating the original verbatim to rewrite it. Token cost grows with the length of the document, and the operation turns bloated. After a few rounds of rewriting, the copies excerpted elsewhere start to drift.

GEML organizes a document into typed blocks, each with a type and a unique id, so a model locates by #id and edits in place. With built-in write validation and .gemlhistory tracking, an agent reads and writes at very low token cost — leaving the precious context window for the actual work.

For people, it is plain text that reads clean; for agents, it is an addressable, verifiable, traceable, revertible "Doc-as-a-Base".

---

GEML is minimal. It is plain text — still clean with no renderer in sight; one block syntax for the whole language; addressable, verifiable, referenceable structure, natively.

Instead of a separate mini-syntax for each kind of content, GEML carries every kind in one container: the typed block. Code is a block. So are tables, diagrams, math, callouts, even metadata — and a run of prose can be one too (=== text), whenever you want it addressable. Extending it later is just as plain. The shape is the same every time, which makes the language easy enough to learn that it's hard to get wrong.

=== code {#hello lang=python}
print("hi")
===

geml get doc.geml '#hello'   # by name, just this block

Blocks have names so the verbs have somewhere to land — the full syntax is in the format in 1 minute.

Contents: What it solves · Why now · What's different · The format in 1 minute · A gift for programmers · Get hands-on · With an LLM · Maturity & versions · The design · Roadmap · Take part · License

What it solves

---

Why the LLM era needs a brand-new text format

Because both the producer and the consumer of a document have changed.

In traditional software engineering, a document was either a static explanation for people to read, or a serialized data file for programs.

Today, people and AI agents collaborate on the same document at high frequency. When the agent becomes the document's "second reader and co-author", the old balance breaks for good:

1. Context is scarce compute: every whole-document read or write burns an agent's limited attention window and reasoning budget;
2. Human–machine collaboration needs an isomorphic carrier: people need to read it at a glance, agents need to read and write it precisely, block by block;
3. Knowledge must have a single source of truth: scattered prompts and copy-pasted Markdown are destined to decay with every iteration.

Yet none of our existing text infrastructure was designed for this scene:

• Markdown (typeset for people): no stable structural blocks, no machine keys. To change one parameter, an agent must read and write the whole text — wasting context budget across multi-turn loops, and inviting drift in both format and meaning.
• JSON / XML (serialized for machines): full of wrapper syntax and structural noise — blocking natural human reading, while quietly eating expensive tokens in long contexts.
• Scratch memory and scattered files (no single source of truth): context is torn across chat history and Markdown copies everywhere; a copy is drift from the moment it is made, and version skew and hallucinated distortion follow.

The answer: "Doc-as-a-Base"

GEML invents no heavy new runtime. It gives plain-text documents one standard set of operational semantics:

A document no longer needs just a format — it needs a set of verbs. GEML keeps plain-text readability and adds deterministic block-level operations.

💡 Deep Dive: If you are interested in the dilemma of engineering documents in the LLM era and why we need to redesign a plain-text format from the ground up, read our full article on the blog: "Why Do We Need a New Text Format in the Era of LLMs?"

---

What's different about GEML

GEML stays small on purpose — the thinking, what it refuses, and what is still open are in how we thought about the design.

The four capabilities were established a chapter ago — addressing, projection, validation, rollback. This chapter is where each format lands against them, and where GEML draws its boundaries.

How other formats compare

Each of the four has mature solutions in its own field; what's unusual is meeting all four in one plain-text format:

Item by item: vs. CommonMark · vs. XML and JSON · a 7-format capability matrix.

Coexisting with Markdown: GEML is the editing source of truth, Markdown is the delivered artifact. Project one way with geml <file> --to md|html and ship .md or .html as before. Collaboration, not lock-in. (Projection is lossy: block ids and table-bound charts don't survive it.)

Don't take the table's word for it — re-run it. This is what I asked the model:

Based on your own experience editing the READMEs just now, describe the command steps you go through on a document (I saw you using grep and such), and whether you cache documents to save tokens — let's compare, and from that see which parts of GEML would actually earn their place.

What came back: what one edit costs and a real day replayed. Paste the question to your own model and see what it tells you. PS: I am still trying to work out whether the upstream chain (who calls this) and the downstream chain (what it calls) that codemap produces can pin down functions and call sites — and change project code — the same way. I will post a report when I have one.

The format in 1 minute

Typed blocks

One shape, every type. A block's basic syntax is === type [attributes] … === (where attributes like {#id .class key=val} are optional) — only the type (and how its body is read) changes:

=== code {lang=python}
print("hi")
===

=== note {.intro}
Parsed prose with *emphasis* and a [[#budget]] reference.
===

=== meta
title = "Budget plan"
===

A run of = (three or more) opens a block; an equal-length run closes it; longer fences nest inside shorter ones. A block that carries an #id can also close with the labeled fence === #id — no fence-length counting, which makes long blocks much harder to get wrong (nesting still requires a longer outer fence: a same-length bare === in the body closes the block early, labeled or not). The type decides how the body is read — raw (verbatim: code, diagram, math, table), flow (parsed prose with inline markup: note, text), or data (one key=val per line: meta); embed carries no body at all — its src= names the block it stands for — and every block may carry an attribute object {#id .class key=val}, where a .class is a semantic label, never a styling hook. The full inline grammar (emphasis, links, [[#id]] auto-references, media, footnotes, inline $math$) is in the spec.

Tables — two bodies, one model

Write a table visually:

=== table {#budget caption="Annual cost"}
| Plan  | Months | Rate |
|-------|-------:|-----:|
| Basic |      1 |   30 |
| Pro   |      2 |   30 |
===

…or as data, with computed columns and a summary row:

=== table {#fy25 format=csv header=1 compute="FY [%.1f] = Q1 + Q2 + Q3 + Q4" summary="Segment = 'Total'; FY [%.1f] = sum(FY)"}
Segment,  Q1, Q2, Q3, Q4
Cloud,     8, 10, 12, 14
Platform,  5,  6,  7,  9
Services,  3,  4,  4,  5
===

Both forms describe the same model. The FY column and Total row are computed at build time:

compute runs + - * / ( ) per row over columns; summary adds a foot row from the aggregates sum / avg / min / max / count (with arithmetic over them, e.g. weighted ratios); a trailing [printf] sets numeric display.

Tables can also pull their data from an external CSV via src="regions.csv".

❓ Up for discussion: should computed columns and the summary row stay? Keep, freeze, or drop — say which.

Math

=== math {#gauss caption="Gaussian integral"}
\int_{-\infty}^{\infty} e^{-x^2} dx = \sqrt{\pi}
===

$$\int_{-\infty}^{\infty} e^{-x^2} dx = \sqrt{\pi}$$

Diagrams & charts — host a DSL, or chart a table

GEML never interprets a diagram body; it routes it to a pluggable renderer (an unknown format is a warning, body preserved):

=== diagram {#flow format=mermaid caption="Review flow"}
graph LR
  A[Draft] --> B{Review} -->|ok| C[Publish]
===

graph LR
  A[Draft] --> B{Review} -->|ok| C[Publish]

A diagram can also chart a table — single source of truth, with the column references checked at build time and no data copied:

=== diagram {format=geml-chart data=#fy25 type=bar x=Segment y=FY}
===

Drawn from the #fy25 table above:

xychart-beta
  title "FY by segment"
  x-axis [Cloud, Platform, Services]
  y-axis "FY"
  bar [44, 27, 16]

Data — a value, not just text

Every block type names what it holds: code a region of code, table a grid, math a formula. data holds a data value, and it is where the data formats live — json (the default) and jsonl today, yaml/toml reserved. Being typed means the body is read, not just displayed: a missing comma fails the build, geml get --json returns the value itself, and a chart can read it directly.

=== data {#log format=jsonl}
{"ts":"09:00","p95":41}
{"ts":"09:10","p95":58}
===

A jsonl body holds one record per line, which a program can blind-append at end-of-file. Records can also stay in their own file: src=ops/latency.jsonl#L900-999 names the file and, optionally, a line window — so the log keeps being appended and tailed as before, while the document is its verified, addressable, chartable view of it.

Embeds — a dynamic reference, not a copy

One block can stand for another: in the same document by src=#id, across documents by src=other.geml#id. An embed is a dynamic lookup of the source at render time — change the source once and every embed follows; delete it and geml check fails the build on the spot.

=== embed {src=#fy25}
===

The body stays empty; the target lives in src=.

Markdown can't show you the projection. To see it live: install the browser extension, open the raw link to sample.geml, and scroll to the Transclusion section — a same-document projection (src=#roadmap), cross-document projections, and even chained resolution (an embed pulls a chart, which itself binds to a table in another file) all render in place: nothing is written there, yet edit the source once and the projection follows.

A gift for programmers — geml-code-graph

To really feel how powerful and flexible a single GEML primitive is, let's try it on a code graph — a familiar but demanding case for programmers: your whole codebase's call graph, written as GEML. geml codemap build lays the call graph out as a tree of GEML documents — every method an #id block, with #calls / #called-by edges both ways. The downstream chain (what a method calls) for troubleshooting, the upstream chain (who calls it) for the blast radius — all visible in a second;

npm i -g @geml/geml
geml codemap build              # --root defaults to . : detect languages -> index -> one merged graph in ./.geml-code-graph/
geml codemap serve              # opens your browser on the graph

[!NOTE] Requirements. Node 22+ for the CLI (npm i -g @geml/geml). Everything below is optional and used only where noted: Joern for non-TS/JS languages in the code graph, and Chrome for the viewer extension.

[!TIP] TS/JS — zero setup: build fetches the scip indexer by itself. Java / C / Python / Go / Kotlin — one extra download, Joern: unzip its release package and pass that folder to build, e.g. --joern ~/joern/joern-cli (--joern C:\joern\joern-cli on Windows), or put it on PATH and skip the flag. Mixed front-end + back-end repo — everything merges into one graph.

geml-code-graph is itself a diagram format — one line embeds it in any GEML document (=== diagram {format=geml-code-graph src=.geml-code-graph/index.geml} ===), and an optional per-commit hook (bundled with the Claude skill) rebuilds it as the code moves, so the graph doesn't drift.

Scale is measured, not promised: on Apache Flink's codebase — 13,585 Java source files, ~81,000 methods, 266,821 call edges — the plain-text data tables still open and query instantly, and you can grep any method name to trace its call chain. Reproduce it yourself: clone apache/flink and run geml codemap build --joern … at its root.

Next — get hands-on now

▶ Try writing GEML in the Playground — edit on the left, rendered live on the right, and the build verdict flips red the moment a reference breaks. No install, nothing to read first.

Then, in the order that suits you:

1. See it render in your browser. Install the extension and open a raw .geml link (the raw file, not the GitHub blob page — that one is HTML): the GEML spec itself (dogfood — the spec is a GEML document, rendered at scale), the showcase (a computed table, four charts, a Mermaid flow, and math), or playground/sample.geml for the interactive code-graph.
2. Run it locally. npm i -g @geml/geml (Node 22+), then geml check a document, or point it at your own repo with geml codemap build.
3. Set up Claude Code — one command. npx -y @geml/geml skill install puts the authoring skill, the CLI and the MCP server in place, user-global, for every project. It edits no settings and installs no hooks. Details.
4. Read the grammar. The full spec (EN / 中文) is normative and short enough to read in a sitting.

Using GEML with an LLM

The goal is one thing: your model edits a block at a time, and verifies — never re-reads and re-emits a whole file to change one paragraph. Getting there takes one step, and which step depends on what you use.

Using Claude Code — run this

npx -y @geml/geml skill install

It installs the authoring skill, the geml CLI and the MCP server, user-global, for every project. No settings.json edits, no hooks; re-run after an upgrade. (Prefer plugins? claude plugin marketplace add geml-spec/geml, then /plugin install geml@geml — same skill, MCP server bundled.)

Using DeepSeek Harness — add this bundle

The same setup, packaged as a dsh bundle — the geml MCP server plus the authoring and code-graph skills:

dsh plugin --profile web add @geml/dsh-plugin   # web = the profile dsh boots by default; use your own profile name if you run another

Listed on dshmarket and awesome-dsh-plugin; source in integrations/dsh-plugin/.

Then say it once in a session, and the project has switched:

This project uses GEML as its base document format; generate other formats from it as needed.

The skill takes it from there. New documents are written as .geml with an id on every section — that id is what later lets one section be replaced instead of the file. Documents that already exist are left where they are: adopting the format is not licence to convert or delete anything. And geml <file> --to md|html produces whatever still has to ship as something else.

Using anything else — paste this, then check the output

A model with no skill to read needs the rules once. Paste the prompt below, and keep geml check as the gate on whatever it writes back — the CLI is npm i -g @geml/geml (Node 22+).

Write the document as GEML: every block is === type [attributes] … === (the format in 1 minute lists the types). Four rules are the ones models get wrong: the closing fence is a = run of the exact opening length, and a body containing === needs a longer outer fence; headings are ATX # only, with no --- frontmatter (metadata is === meta); every #id is unique and every reference ([[#id]], [text](#id), [^id], data=#id) must resolve; there is no raw HTML. The normative spec is GEML-spec.md.

What it will do with it

geml list   doc.geml                                     # CALL FIRST: every block, its address, kind, lines
geml find   "words" doc.geml                             # search block content -> an address, not a line number
geml get    doc.geml '#hello'                            # read ONE block (a heading id = its whole section)
geml get    doc.geml '#hello' --intro                    # a section cuts three ways: --head | --intro | --body
geml set    doc.geml '#license' --in template.geml#mit   # replace that block, forking another
geml add    doc.geml --after '#intro' --in snippet.geml  # insert a fragment (keeps its own ids)
geml revert doc.geml '#plan' --rev -1                    # roll ONE block back
geml check  doc.geml                                     # validate only: diagnostics + exit code

Any section cuts three ways, on get and set alike: --head is the heading line, --intro what it says before its first subheading, --body everything under it — so --body always contains --intro, and equals it when there is no subheading. A section's opening can be edited without pulling its subsections into context.

Every mutation is re-parsed before it writes and refused if it would break the document — which is what makes editing unattended safe. The rest of the verbs (delete, rename, history, --to md|html|geml conversion, addressing a block by type or content hash) are in the parser README.

MCP Server

A standard Model Context Protocol server ships with the package, so your agent edits one block at a time instead of rewriting whole files. It runs locally on Windows, macOS, and Linux; --root is the directory holding your .geml files.

Claude Code / any CLI client — one command:

claude mcp add geml -- npx -y @geml/geml@latest mcp --root /absolute/path/to/your/docs

Claude Desktop — add to claude_desktop_config.json:

{
  "mcpServers": {
    "geml": {
      "command": "npx",
      "args": [
        "-y",
        "@geml/geml@latest",
        "mcp",
        "--root",
        "/absolute/path/to/your/docs"
      ]
    }
  }
}

Then just ask for the change you want — "fix the Q3 row in the FY26 table" — and the agent addresses that one block. You never learn a tool name: each mirrors a CLI verb (geml set → geml_set), so one vocabulary covers the terminal and the agent.

Two guarantees make this better than letting a model rewrite the file: a write is parsed before it reaches disk and refused with its diagnostics if it would break the document, and every write first records a .gemlhistory revision — so a bad edit is both prevented and undoable (geml_revert restores one block, the rest of the file byte-identical). Paths stay confined to --root, which a client cannot widen.

Point --root at a repository that has a code graph (geml codemap build) and the same server also answers "who calls this" — four read-only geml_codemap_* tools, one client entry instead of two. Every tool and option: docs/mcp-guide.md.

Ecosystem and maturity

GEML is a small, young spec — but a stable one: 1.0 is released and usable for real documents (this repo's own spec is one), with a strict conformance suite, a reference implementation that passes it (versioned independently of the spec), and an open proposal process.

Both specs are bilingual:

Versions and compatibility

• Self-hosting — GEML-spec.geml is the specification written in GEML, required to parse clean on every test run.
• A conformance suite is what holds separate implementations compatible.
• A reference implementation of the parser. 1,200+ unit tests today, plus the conformance corpus, round-trip serialization and end-to-end CLI runs, with coverage CI-gated at ≥95% lines / statements / functions / branches.
• Forward compatibility is in the grammar. A processor must degrade gracefully on constructs it does not recognize (spec §8.2), which is why adding a block type or a diagram format is not a breaking change. The type registry is open: an unregistered type name should contain a hyphen (acme-invoice), leaving hyphen-free names to future versions of the spec (§8.5).
• Claiming conformance. An implementation may call itself conformant to GEML 1.0 once it reproduces the conformance suite case for case (§8.5). No permission needed, and no sign-off from this repo.
• On the wire. Extension .geml (version sidecar .gemlhistory), media type text/geml, or text/vnd.geml where a registered type is required — text/geml is not registered with IANA yet.
• A fragment identifier on a .geml URL names the block bearing that id (§0.6) — which is not what #tag means on an HTML page.

How we thought about the design

What the design follows

It is plain text meant for people to read. Fully readable with no renderer — which is why there is no raw-HTML escape hatch, and why a style may never change what a document says.

One primitive, a few models. Every kind of content is the same typed block; extending the format means registering a type, not inventing syntax. The type says what it becomes: meta is key–value shared across the document, code is a region of code at a location, data is a data value, table is a grid waiting to be worked, diagram is a hosted external DSL, embed is a view onto a source of content

A reference is a window, not a navigation. An HTML link navigates: the target is not in the document you are holding, so people copy it in anyway. What is being designed out is not dead links; it is the incentive to copy. Cost: rendering may need to read several files, and must degrade gracefully when it can't.

Prefer subtraction. Where a rule breeds edge cases, the feature goes rather than the edge cases getting specified: no underscore emphasis, no setext headings, no indented code blocks, no raw HTML. The ambiguity is deleted at the source instead of enumerated in test cases. Cost: some things you can write in Markdown you cannot write here.

No broken windows. Markdown's ethos is never to fail — render something. GEML's is the opposite: verified at build time rather than tolerated at render time. A dangling #id is an error with a non-zero exit. Stable ids, geml check and the diagnostic catalogue all follow from that one decision. Cost: a document that "looks fine" can fail your build.

A sidecar travels with the document without getting into it. The .geml file is the source of content and stays deliberately small. Anything else is not pushed into it but points back at it — a version history in .gemlhistory, say — and deleting that leaves the document perfectly valid. Cost: a convention, explicit or implied, and two files that travel together.

The command line is built for an agent. The fewest verbs that cover everything, kept orthogonal, with pipeable input and output and options that stay consistent across them.

What it therefore refuses

Roadmap

• The GEML 1.0 spec (core + history extension), in English and Chinese, with a conformance suite
• Reference implementation @geml/geml: parser, CLI, block-level .gemlhistory tracking
• Official MCP server (geml mcp) for Claude Code, Cursor, Codex and other MCP hosts
• codemap — a whole codebase's call graph, written as GEML
• Ecosystem integrations: VS Code highlighting and reference checking, tree-sitter, Obsidian, the browser viewer, a GitHub Action, LangChain / LlamaIndex, the Claude Code plugin, the Codex plugin, the DeepSeek Harness plugin
• The VS Code extension on the Marketplace
• Parsers in other languages (Rust / Python) — the spec and the conformance suite are public, so community implementations are welcome; we are glad to help line them up

---

Take part

GEML is 1.0, but "stable" means the rules already there won't shift under you, not that the design is settled. There is exactly one implementation so far, and one set of opinions behind the spec. Your thinking can still change the spec itself. If you want a hand in it:

Come argue about these:

• Should the format keep computed columns and summary rows?
• If styling is supported, how should it be designed?
• Is the GEML history file a made-up need?
• --view reads through an embed. Flag, or its own verb?

Or claim a piece:

• Write a second implementation of the spec — a new GEML parser in whatever language you like (how to write a parser)
• Finding the places where the spec is ambiguous is itself the contribution, whether or not that parser ever ships.

Or propose something new:

• A GEP: the proposal, the spec edit and the conformance cases land together (process)

Or put it to use:

Three files to read first: GOVERNANCE.md for how decisions get made, CONTRIBUTING.md for how to send work, and CODE_OF_CONDUCT.md for the one rule about people — disagree with the design as sharply as you like, not with the person.

Repository layout

spec/                  Core spec + .gemlhistory extension as .md (EN / 中文), the
                       CC-BY spec license, and proposals/ (GEPs)
spec/in_geml_format/   The dogfood: those same specs written in GEML, with their
                       .gemlhistory sidecars
geml-parser/           Reference parser, renderer, CLI + codemap toolkit (TypeScript, Node 22)
integrations/          Everywhere GEML plugs in: geml-viewer (browser extension),
                       geml-check-action (CI), vscode, obsidian, tree-sitter (brief)
playground/            In-browser playground (+ a live geml-code-graph of this repo)
docs/                  Guides, design notes, comparisons/ (COMPARISON + vs-CommonMark +
                       vs-XML-and-JSON), assets (logos, used by the Pages site below),
                       and an example .geml to render
.claude/skills/        Claude skills: GEML authoring, and the code graph
.github/               CI + geml-check workflows, MCP registry publish, and issue
                       templates (bug, GEP, new implementation)
site/                  The geml-spec.github.io/geml Pages site: a project homepage
                       (index.md) plus a Jekyll blog (blog/, posts in _posts/) —
                       the long-form "why a new format" article (EN / 中文) lives
                       there as its first post. `cd site && bundle exec jekyll
                       serve` builds it locally; .github/workflows/pages.yml
                       builds and deploys it (grafting in playground/ as static
                       output) on push to main.

License & governance

Code is MIT (LICENSE): everything in this repository — geml-parser/, all of integrations/, playground/, .claude/skills/, the GEPs in spec/proposals/ — except the specification documents.

The specification documents are CC-BY-4.0 (LICENSE-spec.md, which lists them exactly): spec/GEML-spec*, spec/GEML-history-spec*, spec/in_geml_format/*, and docs/comparisons/COMPARISON*. A spec is not software, so anyone may build a conformant implementation without permission — and call it conformant to GEML 1.0 once it passes the conformance suite.

Using the name. You need no permission to implement GEML, to name an implementation after the format (geml-rs, pygeml, a geml package on your language's registry), or to state that your tool reads and writes GEML. Two requests, neither of them a legal restriction: call an implementation conformant to GEML 1.0 only once it passes the conformance suite, and don't imply that this project wrote, endorses, or maintains it. Attribution for the specification text itself is what CC-BY-4.0 already asks for.