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/dreamlonglll/mini-term/claude-mdgit clone --depth 1 https://github.com/dreamlonglll/mini-termWrote 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/dreamlonglll/mini-term/claude-md)<a href="https://agentmods.dev/instructions/dreamlonglll/mini-term/claude-md"><img src="https://agentmods.dev/badge/instructions/dreamlonglll/mini-term/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 | $0.03956 | $0.03956 |
| Opus 5 | $0.01978 | $0.01978 |
| Sonnet 5 | $0.00791 | $0.00791 |
| Haiku 4.5 | $0.00396 | $0.00396 |
Grade B, and why
mini-term CLAUDE.md scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
- Grok 的 hook 接入与另外两家有两处结构性差异，改动前先看 `mt-ai::hook_registry::register_grok_hooks` 的注释：① grok 默认还会扫描 `~/.claude/settings.json` 的 hooks（Claude 兼容层），同一事件会来两趟，sidecar 靠 `GROK_SESSION_ID` + 是否带 argv 丢弃兼容层那趟（只注册了 Claude 的用户必须放行— How it starts
The opening of the file, as written. The whole thing — 145 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
Project Overview
mini-term — GPUI 原生桌面终端管理器，支持多项目、多标签、分屏布局，并能感知 AI 进程（Claude/Codex/Grok 等）状态；带移动端中转镜像与 SSH 远程项目能力。
- UI/渲染: gpui 0.2.x（Zed 官方，crates.io 版）+ gpui-component（Resizable/Modal/Input/Tree 等）
- 终端: alacritty_terminal（VT 状态机，进程内直喂，无 IPC）+ portable-pty
- 发布形态: Windows x64 NSIS 安装包（
scripts/windows-installer.nsi，包内平铺 exe + 三个 sidecar + portable-conpty，全部「与 exe 同目录」）+ macOS dmg + Linux deb/tar.gz - 历史: 项目最初是 Tauri v2 + React 实现，v1.0.0-beta 后整体删除切换到 GPUI 原生版；找旧实现看 git 历史（合并点
236d5c1）
开发命令
# 首次/换版本后：构建三个 sidecar 并连同便携 ConPTY 就位到 target/debug/
node scripts/stage-sidecars.mjs
# 启动开发实例（⚠️ 与装机版并跑时必须隔离数据目录）
MT_APP_DATA_DIR="$LOCALAPPDATA/mini-term-gpui-dev" cargo run -p mt-app
# 全工作区测试（27 个目标 1500+ 例）
cargo test --workspace
# 中转服务端协议边界测试 / sidecar 工作区
cd relay-server && cargo test
cargo build --manifest-path sidecars/Cargo.toml
# 移动端 PWA
cd mobile && npm run build
# 改文案后重新生成 i18n 字典
node crates/mt-i18n/tools/gen_from_ts.mjs
- ⚠️ 禁跑
cargo fmt：本仓 HEAD 非 rustfmt-clean，全仓 fmt 会重排几十个文件淹没 diff。 - ⚠️ GPUI dev 实例运行中时
cargo test -p mt-app会卡在「无法替换 target/debug/mini-term.exe」——先关实例，或cargo test --no-run --message-format=json取出测试二进制直接执行。
架构说明
工作区布局
| 目录 | 说明 |
|---|---|
crates/ |
主工作区（根 Cargo.toml，members = crates/*） |
sidecars/ |
sidecar 二进制独立工作区（miniterm-hook / mt-ssh-mcp / mt-ssh-cli）。版本号自成语义——daemon 换代靠它判断，不跟随主程序发版，故不并入根 workspace。产物由 scripts/stage-sidecars.mjs 就位到主程序 exe 同目录 |
relay-server/ |
移动端中转服务（axum），其 protocol crate 被 mt-relay 跨工作区 path 依赖 |
mobile/ |
移动端 PWA（React + TS + Vite），产物由中转托管 |
crates/ 各 crate 职责
| crate | 职责 |
|---|---|
mt-app |
GPUI 应用壳：Workspace 组件树、AppStore 全局状态、SplitNode 布局树、各面板/弹窗/托盘/标题栏。组件树图见 main.rs 模块注释 |
mt-ui |
GPUI 渲染层：终端 view/element、主题桥。不含业务逻辑 |
mt-terminal |
VT 状态机 + grid 模型（alacritty_terminal 封装）。不依赖 gpui |
mt-pty |
PTY 生命周期（spawn/read/write/resize/kill）+ 便携 ConPTY 预载（conpty.rs，从 exe 旁 portable-conpty/ LoadLibrary 预载） |
mt-ai |
AI 感知：hook server（权威）、hook 注册（hook_registry.rs）、输入检测降级（detect.rs）、状态判定（monitor.rs/perception.rs）、会话记录读取（sessions.rs） |
mt-project |
文件树、目录监听、搜索、Git（git2，vendored-openssl 必须保留）、外部编辑器、WSL 发行版枚举 |
mt-config |
配置持久化(config.db,rusqlite)与主题包。不依赖 gpui。config.json 已退化成给 sidecar 读的 SSH 投影(见下节);界面布局另见 mt-layout |
mt-layout |
界面布局持久化(layout.db,rusqlite):三栏比例 / 每项目分屏树 / 窗口几何。分屏树整棵存 JSON 不拆关系表,理由见模块注释 |
mt-i18n |
双语文案层。字典源头是 locales/*.ts（TS 对象字面量，随 Tauri 版下线迁入），src/dict.rs 由 tools/gen_from_ts.mjs 生成——禁止手改 dict.rs，改文案改 locales 后重跑生成器，tests/consistency.rs 的对账常量随之更新 |
mt-relay |
移动端中转桌面侧：出站 WSS 长连、配对、项目快照/增量、对话镜像（mirror.rs）、移动端指令写穿 |
mt-ssh |
共享 SSH 通信层（russh 持久会话池 + SFTP 原语），主程序与 sidecar 共用 |
mt-usage |
用量统计：会话轮次解析 / SQLite 账本 / 聚合 / 计价 |
mt-core |
叶子共享库（WSL UNC 解析 / SSH 提示扫描 / 原子写等）。⚠️ 依赖方向铁律：只依赖 serde/serde_json/dirs，绝不反向依赖上层 crate——它同时被三个 sidecar 与 mt-ssh 链接 |
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 · 145 lines · 3,956 tokens per session scan B 88aa42824e3a
mini-term CLAUDE.md is an instructions file published in the GitHub repository dreamlonglll/mini-term (217 stars, last pushed 5d ago), licensed MIT. It adds 3,956 tokens to every session, about $0.0198 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
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).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
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.
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.