What is cosmodrome CLAUDE.md?

Repository instructions for Cosmodrome, a native macOS terminal emulator for running multiple AI coding agents across projects. They describe its architecture, performance targets, product boundaries, and development principles.

Understanding the codebase, making performance-sensitive changes, maintaining terminal and agent-state features, and following the project's Swift, AppKit, Metal, and architecture conventions.

They help contributors preserve the app's focus as a terminal and avoid changes that could slow it down or turn it into an IDE or project manager.

How rinaldofesta/cosmodrome configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to rinaldofesta/cosmodrome. 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/rinaldofesta/cosmodrome/main/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/rinaldofesta/cosmodrome

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/rinaldofesta/cosmodrome/claude-md/github.svg)](https://agentmods.dev/instructions/rinaldofesta/cosmodrome/claude-md)

Your own site

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

Per session 2,960 This file is loaded in full into every session.

When invoked 2,960 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	$0.02960	$0.02960
Opus 5.5	$0.01184	$0.01184
Sonnet 5.5	$0.00592	$0.00592
Haiku 4.5	$0.00296	$0.00296

Measured 2mo ago against content hash b7fab58cc1c7, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

cosmodrome 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 2mo 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 · 231 lines

How it starts

The opening of the file, as written. The whole thing — 231 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md — Cosmodrome

What is Cosmodrome

A native macOS terminal emulator for developers who run multiple AI agents (Claude Code, Aider, Codex) in parallel across multiple projects. One window, all projects visible, agent state at a glance.

Built with Swift + AppKit + Metal + libghostty-vt. Not Electron. Not Tauri. Not a wrapper around tmux.

Project Philosophy

We are a terminal. Not an IDE, not an editor, not a file manager. We render terminal output, manage PTY processes, and detect AI agent state. That's it.

Performance is non-negotiable. Every frame under 4ms. Every keystroke under 5ms to screen. If a change makes it slower, revert it. Profile before and after. No exceptions.

Simplicity over cleverness. Two layout modes (grid/focus), not four. Three agent states the user cares about (working/needsInput/error), not seven. One config format (YAML), not two. One Metal view, not N.

Observe, never orchestrate. We watch what agents do and surface it to the developer. We never send input to agents, never auto-trigger actions, never control agent behavior. Read-only observation, always.

Ship, then polish. Phase 0 is a single terminal that renders. Phase 1 adds projects and agents. If it doesn't serve Phase 0-1, it doesn't belong in the codebase yet.

Architecture Overview

Read ARCHITECTURE.md for the full picture. Key points:

Terminal backend: TerminalBackend protocol. Current implementation: SwiftTerm (pure Swift, ships now). Future: libghostty-vt (when C API stabilizes).
Rendering: Single MTKView for the entire content area. Viewport scissoring per session. Shared glyph atlas. One command buffer per frame.
I/O: Single thread with kqueue multiplexing all PTY file descriptors. Agent detection runs inline on output arrival, not on a separate polling thread.
State: @Observable objects. No event bus, no Combine, no NotificationCenter for domain events. Direct property mutation, Observation framework propagates to UI.
Threading: 2 threads total — main (UI + Metal) and I/O (kqueue + VT parsing + agent detection). GCD for background work (config watching, notifications).

Read the full file on GitHub · 231 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.

2mo ago First seen · 231 lines · 2,960 tokens per session scan A b7fab58cc1c7

Subscribe to this mod's changes

cosmodrome CLAUDE.md is an instructions file published in the GitHub repository rinaldofesta/cosmodrome (24 stars, last pushed 7mo ago), licensed MIT. It adds 2,960 tokens to every session, about $0.0118 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.

cosmodrome CLAUDE.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
cosmodrome CLAUDE.mdrinaldofesta/cosmodrome	Instructions file	24	7mo ago	MIT	A	2,960	Claude Code
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	today	MIT	A	7,461	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	5,649	GitHub Copilot
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	today	MIT	A	4,680	Codex, OpenCode
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	3,093	GitHub Copilot
vscode model-prompts.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	2,022	GitHub Copilot

cosmodrome: Instructions file for Claude Code