What is rabbithole AGENTS.md?

Repository instructions for developers working on Rabbithole, an MCP server that opens a branching-document canvas in a web browser. They describe the project structure, development commands, debugging, conventions, and tests.

Use them when developing, debugging, testing, or changing the browser interface, document engine, server connections, storage, or generated interface bundles.

They provide the project-specific context an agent needs before changing code. This reduces the risk of editing the wrong part of the repository or missing required build and test steps.

How shlokkhemani/rabbithole configures Codex and OpenCode

Reuse

Borrowing it

Nothing to install: this file belongs to shlokkhemani/rabbithole. 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/shlokkhemani/rabbithole/main/AGENTS.md

Clone the repo

git clone --depth 1 https://github.com/shlokkhemani/rabbithole

Made for: Codex, OpenCode.

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/shlokkhemani/rabbithole/agents-md/github.svg)](https://agentmods.dev/instructions/shlokkhemani/rabbithole/agents-md)

Your own site

<a href="https://agentmods.dev/instructions/shlokkhemani/rabbithole/agents-md"><img src="https://agentmods.dev/badge/instructions/shlokkhemani/rabbithole/agents-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/shlokkhemani/rabbithole/agents-md"><img src="https://agentmods.dev/badge/instructions/shlokkhemani/rabbithole/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 619 This file is loaded in full into every session.

When invoked 619 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	$0.00619	$0.00619
Opus 5.5	$0.00248	$0.00248
Sonnet 5.5	$0.00124	$0.00124
Haiku 4.5	$0.00062	$0.00062

Measured 1mo ago against content hash a8323f53ec29, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

rabbithole AGENTS.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 1mo 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.

AGENTS.md · 53 lines

How it starts

The opening of the file, as written. The whole thing — 53 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Notes for agents working in this repo

If you were sent here to install Rabbithole for a user, stop — you don't need to clone or build anything. Follow the Quick start in README.md (one claude mcp add / codex mcp add line). This file is for agents developing the repo.

What this is

An MCP server (stdio) that opens a branching-document canvas in the browser. Plain ES modules, a small esbuild-based browser build, and script-driven tests.

bin/mcp-server.js — entry; just imports src/node/mcp/server.js
src/core/ — host-independent document engine, renderer, artifacts, and contracts
src/ui/ — browser runtime shared by live pages and frozen snapshots
src/node/ — MCP wiring (server name rabbithole), filesystem storage, sessions, local HTTP/SSE transport, and Node PDF ingestion
src/web/ — static BYOK browser host, provider adapters, and IndexedDB store
src/core/html/ — shared self-contained shell, tokens, and stylesheet source
src/core/html/icons.js — canonical repository for all product-owned SVG icons and brand marks
dist/ — ignored live and frozen UI bundles; built during install and packaging
test/ — capability-oriented suites documented in docs/testing.md
website/public/ — live public assets copied by build:publish

Run / debug

npm install
RABBITHOLE_NO_BROWSER=1 node bin/mcp-server.js   # speaks MCP on stdio
npm run build                                    # regenerate package and web bundles
npm test                                         # deterministic default suite

Storage is JSON files under ~/.rabbithole/ (RABBITHOLE_DIR overrides). Logs go to stderr — stdout is reserved for the MCP protocol; never print to stdout.

Conventions

The product name is Rabbithole — one word, no space, in all copy.
Node ≥ 18, ES modules everywhere.
The canvas page must stay fully self-contained (one HTML response, no external assets) — that constraint is load-bearing for export/snapshots.
stdout is reserved for MCP protocol messages; application logs go to stderr.
Preserve old .rabbithole files and snapshots according to docs/compatibility.md; future formats must fail clearly rather than truncate.
Put every product-owned SVG icon or brand mark in src/core/html/icons.js and render it with iconSvg(). Do not add inline icon geometry to shell, UI, web, settings, or website files. Structural/document SVG (for example the canvas edge layer or user-authored content) is not an icon and remains at its owning trust boundary.

Read the full file on GitHub · 53 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.

1mo ago Changed · +3 tokens per session a8323f53ec29
2mo ago First seen · 53 lines · 616 tokens per session scan A 46eb67800d1c

Subscribe to this mod's changes

rabbithole AGENTS.md is an instructions file published in the GitHub repository shlokkhemani/rabbithole (327 stars, last pushed 10d ago), licensed MIT. It adds 619 tokens to every session, about $0.0025 per session on Opus 5.5. 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 2026-08-30.

rabbithole AGENTS.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
rabbithole AGENTS.mdshlokkhemani/rabbithole	Instructions file	327	10d ago	MIT	A	619	Codex, OpenCode
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	yesterday	MIT	A	7,461	Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	yesterday	MIT	A	4,680	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	5,516	GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	yesterday	MIT	A	3,093	GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness	Instructions file	274k	2d ago	MIT	A	4,313	Codex, OpenCode

rabbithole: Instructions file for Codex