forestui CLAUDE.md

A project instruction file for AI coding assistants working on ForestUI. It describes the project, its technology, its file structure, and a strict rule not to interact with the user's live tmux session.

In plain words
What is it for?
Use it as operating guidance when an AI assistant edits or tests the ForestUI codebase.
Why use it?
It prevents an assistant from disturbing active editors, terminals, or conversations while working in the repository. It also provides project-specific context before changes are made.

Instructions file

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add instructions/flipbit03/forestui/claude-md
Clone the repo
git clone --depth 1 https://github.com/flipbit03/forestui
Per session 8,041 This file is loaded in full into every session.
When invoked 8,041 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
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.

ModelPer sessionOnce invoked
Fable 5 $0.08041 $0.08041
Opus 5 $0.04020 $0.04020
Sonnet 5 $0.01608 $0.01608
Haiku 4.5 $0.00804 $0.00804

Measured 2d ago against content hash 88fa50304988, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

forestui 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 2d 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 · 628 lines

How it starts

The opening of the file, as written. The whole thing — 628 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md - AI Development Guidelines

This file provides context for AI assistants (like Claude) working on forestui.

Never touch the user's live tmux

This rule applies to every task in this repo, not just testing.

Whoever works on forestui is, almost by definition, running forestui. That is what makes this different from an ordinary "be careful" note: a live tmux server is up right now, holding their actual work — editors mid-edit, terminals mid-command, Claude sessions mid-conversation, laid out the way they left them. It is not a risk to weigh against convenience; it is the default state of this repo's development environment, and a bare tmux command from a shell attaches straight to it.

The session running this task is very likely one of those windows — forestui opens Claude sessions in tmux windows, and this is a Claude session. A stray rename-window, kill-window or kill-server can destroy the user's work, up to and including the conversation that issued the command.

  • Never run a bare tmux command — not to look, not to list, not "just once". tmux list-sessions is as dangerous as tmux kill-server, because it is the step that convinces you the next call is safe.
  • Drive tmux only through tu, with TMUX_TMPDIR pointed at a throwaway directory. The test-forestui skill has the exact invocation.
  • Exploring tmux behaviour is not an exception. Learning what a tmux command does, probing options or hooks, reproducing something from the docs — all of it goes on an isolated server under tu. A separate -L socket looks like it solves this, and it does not: it relies on never once forgetting the flag, against a default that reaches the user's live session.

The test-forestui skill carries the full harness recipe, and its description asks to be loaded before any tmux command, not only before testing — because tasks that do not look like testing are exactly where this gets broken: a spike, doc research, "just checking how tmux options work". The rule is repeated here because this file is always loaded and a skill is not.

Read the full file on GitHub · 628 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.

  1. 2d ago First seen · 628 lines · 8,041 tokens per session scan A 88fa50304988

Subscribe to this mod's changes

forestui CLAUDE.md is an instructions file published in the GitHub repository flipbit03/forestui (25 stars, last pushed 6d ago), licensed MIT. It adds 8,041 tokens to every session, about $0.0402 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 2026-08-30.