By geminiicode
Open a markdown file in a Google-Docs-style comment editor and let Claude address your comments. Wraps the `mdc` CLI.
A desktop markdown viewer with Google-Docs-style commenting and an AI review loop powered by your own Claude Code.
Read a markdown file, highlight text, and leave comments in the margin. A watcher notices each new comment and has Claude revise the document to address it — replying to every comment as it goes — and the editor live-reloads to show the changes. Comments are stored in a hidden sidecar next to each file, so they travel with the document and play nicely with git.
highlight + comment ──▶ mdc revises via your Claude Code ──▶ editor live-reloads
▲ │
└──────────────────────── you review ◀─────────────────────────┘
The editor is the human side; a small CLI (mdc) is the AI side; a hidden
.<file>.md.comments.json sidecar is the shared state both read and write. The
revision shells out to your installed claude CLI (claude -p) — so there
are no API keys to manage and no separate billing; it uses your existing Claude
Code session.
claude CLI being on your PATH. See
https://code.claude.com/docs. Verify with claude --version.Installs straight from GitHub — no registry involved:
npm install -g github:geminiicode/markdown-commenter
Electron ships as a dependency, so the editor launches with no extra build step.
The whole loop in one command:
mdc review notes.md # a single file
mdc review ./docs # or a whole folder
This opens the editor and starts a watcher that auto-revises whenever you add a comment:
Closing the window (or Ctrl-C) stops the watcher.
If you'd rather drive the pieces yourself:
mdc address <file.md> # one-shot: revise the doc to address open comments
mdc watch <file.md> # just the auto-address watcher (no editor)
mdc list <file.md> # print comments and their statuses
You can also launch just the editor with npm start and open files from the UI.
This repo doubles as a Claude Code plugin. Add it as a marketplace and install:
/plugin marketplace add geminiicode/markdown-commenter
/plugin install markdown-commenter
Then run the review loop from a Claude Code session (plugin skills are namespaced by plugin name):
/markdown-commenter:review-md notes.md
The plugin handles the mdc prerequisite itself: on the first session after
install it pre-installs the CLI in the background, and if that hasn't completed
by the time you invoke the skill, the skill offers to run the install
interactively. Installing via npm yourself (above) also works and skips both.
The review loop runs Claude with edit access to the document's directory
(--permission-mode acceptEdits), and both the document and your comments are
part of the prompt. That means reviewing a markdown file is effectively
granting its author edit access to files in that directory — a malicious
document could instruct Claude to make unwanted edits. Claude is restricted to
Read,Edit (not Write), so it edits existing files in place but can't create
new ones, which bounds the blast radius.
Note that pointing mdc review (or mdc watch) at a folder runs Claude
automatically and unattended on every document in the tree as you comment, so
the trust requirement extends to every file under that folder. Only run the
review loop on documents you trust.
Comments live in a hidden sidecar beside each file: notes.md →
.notes.md.comments.json. Each comment carries the quoted text it anchors to,
a status (open → addressed → resolved), and a thread of replies. The
schema is plain JSON — the editor and CLI are just two clients of it.
MIT — see LICENSE.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub geminiicode/markdown-commenter --plugin markdown-commenterClaude + Obsidian knowledge companion. Sets up a persistent, compounding wiki vault (Karpathy's LLM Wiki pattern). v1.7 "Compound Vault" + v1.8 methodology modes close 5 of 5 priority gaps from the May 2026 compass artifact. Ships: substrate alignment with kepano/obsidian-skills, default Obsidian CLI transport, hybrid retrieval (contextual prefix + BM25 + cosine rerank per Anthropic's Sept 2024 research), per-file advisory locking for multi-writer safety, pre-commit verifier agent, AND methodology modes (LYT / PARA / Zettelkasten / Generic) for first-class organizational support no other Claude+Obsidian competitor offers. v1.7.x audit closure: every BLOCKER + HIGH + MEDIUM + LOW finding from the v1.7.0 audit is CLOSED or DEFERRED-with-rationale. Optional DragonScale Memory extension (log folds, deterministic addresses, semantic tiling lint, boundary-first autoresearch).
Complete AI coding workflow system. Self-correcting memory + persistent FTS5-indexed research wikis + auto-research loop + multi-LLM council on a single SQLite store. 33 skills, 8 agents, 22 commands, 37 hook scripts across 24 events. Cross-agent via SkillKit.
Make your AI agent code with your project's architecture, rules, and decisions.