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.
npx agentmods add instructions/far-se/tabame/claude-mdgit clone --depth 1 https://github.com/Far-Se/tabameWrote 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.
[](https://agentmods.dev/instructions/far-se/tabame/claude-md)<a href="https://agentmods.dev/instructions/far-se/tabame/claude-md"><img src="https://agentmods.dev/badge/instructions/far-se/tabame/claude-md.svg" alt="Measured on agentmods" height="20"></a>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 | $0.01271 | $0.01271 |
| Opus 5 | $0.00635 | $0.00635 |
| Sonnet 5 | $0.00254 | $0.00254 |
| Haiku 4.5 | $0.00127 | $0.00127 |
Grade A, and why
tabame 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 4d 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.
How it starts
The opening of the file, as written. The whole thing — 46 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 with code in this repository.
What this is
Tabame is a Windows-only Flutter desktop app that replaces the taskbar with a "QuickMenu" — a hotkey-summoned popup with audio/media controls, pinned apps, a window switcher, an app launcher, bookmarks, timers, clipboard history, color picker, screenshot tools, and many other small utilities. It's a single Win32 desktop app built with window_manager for frameless/transparent windows and a custom tabamewin32 Flutter plugin for native Win32 interop (global hooks, media session, hotkeys).
Commands
- Run/debug: run
flutter run -d windows. - Build release:
flutter build windows. - Lint:
flutter analyze(usesanalysis_options.yaml, based onpackage:lints/recommended.yamlwithalways_specify_types,prefer_relative_imports, and const-preferring rules enabled). - Tests:
flutter test(all),flutter test test/launcher_core_test.dart(single file). - The
tabamewin32directory is a separate local Flutter plugin package (path dependency inpubspec.yaml) containing the native Win32 glue — edit it like a normal Flutter plugin (Dart intabamewin32/lib, C++ intabamewin32/windows).
Architecture
Startup flow
lib/main.dart parses CLI arguments first — Tabame relaunches itself as a separate process for several auxiliary windows (-spotlight, -editor, -screenCapture, -screenRecording, -screenDraw, -colorPicker, -msgbox, -run), each its own minimal entry point in lib/pages/. If none match, it goes through AppStartup (lib/logic/app_startup.dart): admin-relaunch check → register services/hooks → window_manager setup (frameless, transparent, always-on-top for QuickMenu; normal window for the Interface/settings page) → runApp(Tabame).
Multi-window-via-single-exe pattern
There isn't a router with multiple windows in one process — each "extra" surface (spotlight search, photo editor, screen capture/draw/recording, color picker, message boxes, run-status) is a new process launch of the same exe with a flag, handled at the top of main(). When adding a new standalone overlay/tool, follow this pattern rather than trying to open a second Flutter window in-process.
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.
- 4d ago First seen · 46 lines · 1,271 tokens per session scan A cd1ce43eb0e6
tabame CLAUDE.md is an instructions file published in the GitHub repository Far-Se/tabame (366 stars, last pushed 3d ago), licensed MIT. It adds 1,271 tokens to every session, about $0.0064 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.
Other instructions, from other repositories
appwrite AGENTS.md
AGENTS.md instructions for appwrite/appwrite, covering appwrite, commands, stack, layout and libraries.
nim_duilib CLAUDE.md
Instructions for rhett-lee/nim_duilib, covering nimduilib - 跨平台 c++ ui 库, 项目概述, 项目结构, 开发模式（xml + c++） and xml 布局文件.
ContextMenuMgr AGENTS.md
AGENTS.md instructions for PLFJY/ContextMenuMgr, covering agents.md, 1. required reading, 2. identify the correct flow before editing, 3. hard rules and user context.
AIUsageTracker AGENTS.md
AGENTS.md instructions for rygel/AIUsageTracker, covering ai usage tracker - monitor guidelines, critical rules, never wipe user settings (critical), never cause data loss in migrations (critical) and never create releases without explicit permission.
AIUsageTracker CLAUDE.md
Claude Code instructions for rygel/AIUsageTracker, covering claude.md, build & test, analyzer rules — do not weaken, architecture and data flow: monitor → main window.
ContextMenuMgr CLAUDE.md
Claude Code instructions for PLFJY/ContextMenuMgr: The main agent entry point for this repository is.