# ZennoPoster: browsers of running tasks

Connects coding agents to active ZennoPoster task browsers via MCP. It allows models to click elements, inspect web pages, and capture screenshots during execution.

- MCP server · `type=mcp` · Claude Code, Codex, Cursor, GitHub Copilot, Windsurf (Devin Desktop), Cline, Zed · ★ 7
- Author: ZennoLab
- License: MIT · policy=link-only · description only; the file stays with the author
- Source: https://github.com/ZennoLab/zennoposter-mcp
- Updated: 2026-10-21
- HTML: https://affpapa.org/en/nevibecode/mcp/zennoposter-browsers-of-running-tasks

## What it does

The `MCP.Instance` server connects external LLM clients directly to browsers running inside ZennoPoster tasks. Through this bridge, an agent can navigate web addresses, inspect DOM nodes, click buttons, and save page screenshots. To hook into task instances instead of the template editor, launch the process with `Target=zennoposter` pointing to port `:5300`.

A running ZennoPoster 7.9.3 or higher is required. The server ships as an autonomous `win-x64` binary, so you do not need an external .NET installation. ZennoPoster reserves ports 6107 to 6113 for its internal chat assistants. For this reason, standalone client setups must run on assigned external ports such as 6209.

Authorization relies on an ApiKey generated in ProjectMaker under the API keys settings. Place this token in `appsettings.json`, pass it via environment variables, or provide it on the command line. Inbound `Authorization` headers are ignored, and the service accepts connections only over loopback.

## Install

### Claude Code (native platform)

**Install**
- The MCP registry has no launch command for it: see the repository README

## When to use it
- click buttons and links in active ZennoPoster task instances
- capture screenshots from running browser automation jobs
- navigate to target pages inside background worker threads
- extract element structures and text from opened tabs

## Example prompts
- Navigate to the target website in the active ZennoPoster task and take a screenshot.
- Click the submit button inside the currently running task browser.
- Inspect the active page DOM and extract the form input fields.

## Similar
- [Scrapling MCP Server](https://affpapa.org/en/nevibecode/mcp/scrapling-mcp-server.md) — MCP server · ★ 89899
- [browser-use](https://affpapa.org/en/nevibecode/mcp/browser-use.md) — MCP server · ★ 119231
- [AIHawk](https://affpapa.org/en/nevibecode/mcp/aihawk.md) — MCP server · ★ 32260
- [Skyvern](https://affpapa.org/en/nevibecode/mcp/skyvern.md) — MCP server · ★ 23283
- [SeleniumBase MCP](https://affpapa.org/en/nevibecode/mcp/seleniumbase-mcp.md) — MCP server · ★ 13181
- [Inbox Zero](https://affpapa.org/en/nevibecode/mcp/inbox-zero.md) — MCP server · ★ 12826

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md