What is marginalia CLAUDE.md?

Instructions for working on Marginalia, a tool that turns Markdown into a browser page where people can click text and leave threaded comments through an MCP connection. It covers the project's purpose, code structure, and checks to run.

Use it when modifying Marginalia's MCP server, comment threads, Markdown rendering, local web server, transcript export, or browser assets.

It gives a coding assistant the project-specific context needed to change the right files and verify that the changes work.

How Jin-HoMLee/marginalia configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to Jin-HoMLee/marginalia. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/Jin-HoMLee/marginalia/main/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/Jin-HoMLee/marginalia

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/jin-homlee/marginalia/claude-md/github.svg)](https://agentmods.dev/instructions/jin-homlee/marginalia/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/jin-homlee/marginalia/claude-md"><img src="https://agentmods.dev/badge/instructions/jin-homlee/marginalia/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/jin-homlee/marginalia/claude-md"><img src="https://agentmods.dev/badge/instructions/jin-homlee/marginalia/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 969 This file is loaded in full into every session.

When invoked 969 The same file — it is already loaded in full.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00969	$0.00969
Opus 5	$0.00485	$0.00485
Sonnet 5	$0.00194	$0.00194
Haiku 4.5	$0.00097	$0.00097

Measured 8d ago against content hash 25a7a1aac2c0, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

marginalia CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

CLAUDE.md · 84 lines

How it starts

The opening of the file, as written. The whole thing — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working in this repository.

What this is

marginalia turns any Markdown into a browser-based, click-to-comment thread over MCP. It is one portable FastMCP-over-stdio server, launched with uvx straight from git, that installs identically across Claude Code, Cline, and opencode. The user clicks an element in the rendered page to comment; the agent receives that comment as a clean, anchored MCP tool_result; the agent's reply renders as an in-page card the user can re-annotate (threaded). The core contract is clean recording — comments flow only through the MCP tool loop, never via a side file.

Architecture

src/marginalia/, entry point marginalia.mcp_server:main:

mcp_server.py — the MCP server + tool loop (start_thread · await_comment long-poll · post_reply · end_thread).
store.py — in-memory thread/comment state.
render.py — Markdown → annotatable HTML page.
http_face.py — the localhost server that serves the page + receives clicks.
export.py — writes the Markdown transcript on end_thread.
assets/ — browser JS/CSS for the click-to-comment page.

⚠️ Verify by running

This project has runtime-only surfaces — a stdio MCP server, browser JS, and a long-poll loop — so static review has a structural blind spot: runtime-only failures slip past careful reading and multiple AI reviewers. A README warm-line --help hang once survived four static passes before being caught by running it. Run it to verify (drive the real tool loop / open the page in a browser). See cerebrum's feedback_review_by_running for the canonical case.

Build / test

python3 -m venv .venv && . .venv/bin/activate
pip install -e ".[dev]"
pytest          # 44 tests

A drop below 44 means a test was lost — investigate, don't paper over.

Configuration / env knobs

MARGINALIA_POLL_S — await_comment long-poll bound, seconds (default 540). Set 20 on opencode (its MCP exec cap is ~30s would kill the default long-poll); ~540 is fine on Claude Code and Cline.
MARGINALIA_THREADS_DIR — base dir for exported transcripts (default ~/.marginalia/threads/).
MARGINALIA_PORT — preferred localhost port (default 8787; falls back to an ephemeral port if busy).

Read the full file on GitHub · 84 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

8d ago First seen · 84 lines · 969 tokens per session scan A 25a7a1aac2c0

Subscribe to this mod's changes

marginalia CLAUDE.md is an instructions file published in the GitHub repository Jin-HoMLee/marginalia (0 stars, last pushed 2mo ago), licensed MIT. It adds 969 tokens to every session, about $0.0048 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Writing and docs

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,296 tokens

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,153 tokens

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

marginalia: Instructions file for Claude Code