ming-14/PTY-Agent
Give your AI agent a real terminal: persistent PTY sessions, prompt-triggered returns, screen snapshots, and TUI/GUI/crash awareness. | 给 Agent 一个真实的终端
Project Overview项目介绍
PTY-Agent is a DSH-native CLI plugin that enables AI agents to interact with pseudo-terminals just like a human user. Unlike traditional agent terminal integrations that only deliver unprocessed raw stdout byte streams, PTY-Agent captures fully rendered terminal screen snapshots, letting AI work with interactive REPLs, debuggers, TUIs, and installers that require user input mid-process. It uses ConPTY on Windows and openpty on Linux, integrates natively with DSH agents, and supports triggers that return control to the agent when a specific regex prompt appears on screen.
This plugin is designed for AI-assisted developers who need to complete complex interactive terminal tasks that most general-purpose agents cannot handle. A common workflow starts with the DSH agent launching a PTY-Agent session for an interactive program like gdb or SSH, then waits for the specified prompt to appear before sending the next sequence of input. It supports persistent sessions, asynchronous non-blocking execution, and YAML-based DAG workflow orchestration for multi-session tasks, letting developers delegate complex end-to-end workflows to their AI assistant.
PTY-Agent currently officially supports Windows 10 and newer, and all modern stable Linux distributions, with macOS support still in active development. To install, you can either download a pre-built release package from the repository’s Releases page, or clone the full source code and run the included build.py script to compile it yourself. Most of the codebase is licensed under the MIT license, but the web static assets and RIME input method code are licensed under AGPL-3.0, and the sandbox security feature is currently only available on Windows.
PTY-Agent 是一款专为 DSH 开发的原生 CLI 插件,核心能力是为 AI 代理提供可交互伪终端环境,让 AI 能像真实用户一样操作终端内的 REPL、调试器、TUI 应用等交互式程序。它基于 Windows ConPTY 和 Linux openpty 实现,可获取渲染后的终端屏幕快照,而非传统方式的原始字节流,支持 DSH 代理直接调用。
开发者可以借助 PTY-Agent 让 AI 代理完成需要交互式终端的开发任务,比如远程 SSH 登录、调试二进制程序、运行交互式安装程序等。AI 可以等待特定提示符出现后再继续下一步,支持持久化会话、异步运行和多会话工作流编排,适合需要 AI 处理复杂交互式终端任务的开发人员。
PTY-Agent 当前支持 Windows 10+ 和 Linux,macOS 支持仍在开发中。安装可直接下载预编译的 Release 包,也可克隆源码后运行 build.py 编译。项目大部分代码采用 MIT 许可证,web 静态资源和 RIME 部分采用 AGPL-3.0 许可证,沙箱功能目前仅在 Windows 可用。
请帮我安装这个 DSH 插件。安装前先完成【兼容性检查 + 安全性检查】,检查通过再动手。
插件:PTY-Agent(ming-14/PTY-Agent)
仓库:https://github.com/ming-14/PTY-Agent
本站详情页:https://www.yhbd.top/plugins/ming-14-pty-agent/
本站登记:类型 client · 归类 原生 DSH 插件 · 许可证 MIT · ⭐ 2 · 最近提交 2026-10-01 · 主语言 Python
按下面顺序执行,每步先把结论告诉我,再进入下一步:
【1 兼容性检查】
① 我这边:DSH 版本、Node 版本、操作系统、当前 profile(web / desktop)。
② 读它的 README、package.json、插件 manifest,列出它要求的 DSH 版本 / Node 版本 / 操作系统 / 外部依赖 / 需要另外先装的运行时。
③ 逐条比对,结论只写「满足 / 不满足 / 未知」三种;不满足的给出可行替代方案。
④ 检查是否和我已装的插件冲突:命令名重复、skill / tool 重名、端口占用、重复注册的 MCP server。
【2 安全性检查】
① 仓库可信度:和上面「本站登记」是否一致;star / fork 数、创建时间、最近提交,是否归档或长期停更。
② 安装脚本:逐行看 package.json 的 preinstall / install / postinstall,以及 install.sh、setup.ps1 之类脚本。出现 curl|bash、下载后直接执行、混淆代码、访问与插件功能无关的域名,立刻停下来告诉我,不要继续装。
③ 依赖:列出新增依赖,标出无人维护、或与知名包拼写近似的可疑包(typosquatting)。
④ 权限与副作用:它会读写哪些目录、访问哪些域名、需要哪些 DSH 权限(filesystem / network / shell / clipboard 等),以及怎么卸载和回滚。
⑤ 如果它要求 sudo / 管理员权限,或权限明显超出功能所需,先停下来问我。
【3 安装】
上面两步没有「不满足」和「高危项」时才执行;用官方推荐方式安装,不要自行提权。
【4 汇报】
用表格输出:检查项 / 结论 / 依据 / 是否需要我决策。拿不准的一律写「未知」并说明要我怎么确认——不要猜,也不要替我决定。
Send this message to DSH in your current session: it verifies compatibility and security first (answering met / not met / unknown item by item) and only installs once everything checks out — it will stop and ask you if it finds a high-risk item. The box scrolls; the copy is the full prompt. CLI install commands may not be accurate across systems, so DSH is the safer route.把上面这条消息直接发给当前会话里的 DSH:它会先核对兼容性与安全性(逐条给「满足 / 不满足 / 未知」),确认没问题再安装,有高危项会停下来问你。框内可滚动,复制到的是完整提示词;安装命令不一定准确,发给 DSH 更稳。
- Only 2 stars - very few users, little community feedback星标只有 2,几乎没人在用,遇到问题缺少社区反馈
DSH walks through these 9 checksDSH 会逐条核对这 9 项
Compatibility兼容性
- DSH, Node, OS and profile requirementsDSH 版本 / Node 版本 / 操作系统 / profile 是否满足要求
- External dependencies and runtimes (Electron / Python / Docker, ...)外部依赖与运行时(Electron / Python / Docker 等)是否齐备
- Conflicts with installed plugins: command names, skill / tool names, ports, duplicate MCP registration与已装插件是否冲突:命令名、skill / tool 重名、端口占用、重复 MCP 注册
Security安全性
- Repo matches the facts registered here; archived or abandoned?仓库是否与页面登记一致,是否归档或长期停更
- Safety of preinstall / install / postinstall and install.sh / setup.ps1preinstall / install / postinstall 与 install.sh、setup.ps1 是否安全
- curl|bash, download-then-execute, obfuscation, unrelated domains → stop immediatelycurl|bash、下载即执行、混淆代码、无关域名 → 立刻停止
- Typosquatting or unmaintained packages among the new dependencies新增依赖里有没有 typosquatting 或无人维护的包
- Requested permissions vs. what the feature actually needs申请了哪些权限、是否超出功能所需(filesystem / network / shell / clipboard)
- Any sudo / admin requirement, plus uninstall and rollback是否要求 sudo / 管理员权限,以及卸载与回滚方式
Anything uncertain must be marked unknown with a note on how to confirm it. This site's signal screen is a static snapshot, not a security audit.拿不准的必须标「未知」并说明要我怎么确认。本站的信号筛查是静态快照,不能替代安全审计。
Or use CLI install (for developers)或使用命令行安装(适合开发者)
CLI Install命令行安装
dsh plugin --profile web add github:ming-14/PTY-Agent
把 ming-14/PTY-Agent 加入你的 DSH 配置(web profile)即可启用。
READMEREADME
Let AI use terminals like a real user.
Drive REPLs, debuggers, TUIs, installers, long-running services, and coding agents — getting what the user actually sees on screen, not just a raw stdout byte stream.
Features
- ✨ Shell · CUI · TUI — use terminals like a human: start sessions, send keys, wait for prompts, read screens
- ✨ Cross-platform — Windows / Linux, macOS coming soon
- ✨ Real-time Web monitoring — watch and control every terminal session, collaborate live
- ✨ Sandbox — restrict AI operations to the workspace, prevent accidental damage
- ✨ Plugin system — extend functionality freely
- AI secondary analysis — pipe long logs, large outputs, even rendered terminal screenshots to another AI
- Sub-agents — launch OpenCode / Claude Code / ... across harnesses, managed centrally
- Workflow orchestration — multi-session DAG choreography
- Remote terminal access — SSH-like experience
Why PTY-Agent?
Traditional agents can't run ssh, gdb, cdb, or any program that asks you a question.
Showing the opening section of the README — the full document lives in the repository以上为 README 开头摘要,完整文档在仓库内 · View the full README on GitHub →在 GitHub 查看完整 README →
robiteame/dsh-session-tree-extension
chiro2001/dsh-oc
shiningsprk-arch/dsh-context-viewer
kevenxz/dsh-desktop
ccch1mneyyy/dsh-TUI
xyTom/coding-tools-mcp