# Paneherd Review VS Code and Cursor extension that runs Claude Code, Codex, Copilot, Gemini, and more in a single agent grid. ## Editor's Score: 72/100 Capability scores 20 because the multi-agent grid, conflict guard, merge queue, and phone remote are genuinely novel features no individual agent CLI provides, but the permission inbox and conflict guard only work for Claude Code sessions, leaving other agents as dumb terminals with status display only. Ease of use scores 19 because the install path is a standard VS Code extension with no account setup, the QR-code phone pairing is three steps, and the free trial removes purchase friction, though the feature gating between free and Pro requires reading documentation to understand. Value scores a full 25: a one-time $2.99 license with offline validation, no subscription, no seats, and a 14-day refund is genuinely exceptional pricing for the category, and the free tier is meaningfully functional. Delivery scores 8 because the VS Code Marketplace listing shows zero user reviews and a very low install count at the time of this review, meaning there is no independent evidence base for real-world reliability; the tool is very new and unproven in production, which is a material risk for a buyer considering it for daily multi-agent workflows. ## Pros - One-time $2.99 Pro license with no subscription, no seats, and offline license validation means the cost never scales against you. - Conflict guard actively intercepts edits when two agents are about to modify the same file, a problem no individual agent CLI solves on its own. - Phone remote works via browser with end-to-end encryption and no app install, letting you answer permission prompts and send prompts from away from your desk. ## Cons - The permission inbox and conflict guard, the two features most likely to justify the tool for multi-agent users, are both Pro-only; the free tier shows status but cannot intercept or answer prompts remotely. - Paneherd only runs inside VS Code 1.95 or later and Cursor; developers on JetBrains IDEs, Neovim, or any other editor have no path to use it. - The VS Code Marketplace listing shows zero reviews and a very low install count, so there is little independent user evidence about real-world reliability or edge-case behavior. ## Pricing Freemium. A free tier is available. - Free: $0 - Pro: $2.99 one-time Pricing reflects what we saw at the time of review (2026-11). Confirm current pricing on the official site. Category: AI Coding & Dev Tools (https://aitoolseekers.com/category/ai-coding-dev-tools) Official site: https://paneherd.com/ Last verified by a human: 2026-11-16 Reviewed by a human at AI Tool Seekers, a hand-reviewed AI tools directory. Every listed tool is verified by a person before publishing. Scores are editorial opinions from a fixed rubric (capability, ease of use, value, delivery on promise) and cannot be bought; there is no pay-to-rank. Directory: https://aitoolseekers.com This review: https://aitoolseekers.com/tools/paneherd Methodology: https://aitoolseekers.com/how-verification-works ## Verdict Paneherd is the right pick for developers already juggling two or more AI coding agents who are tired of hunting through scattered terminals for the one waiting on a permission prompt. The permission inbox and conflict guard are genuinely useful, but both are locked behind Pro, and the extension only runs inside VS Code or Cursor, so JetBrains and Neovim users are out. If you use a single agent and want a full AI-first editor rather than an orchestration layer, Cursor is the better fit. Best for: Developers running multiple AI coding agents simultaneously in VS Code or Cursor. Not for: Single-agent users, JetBrains or Neovim users, or anyone wanting a fully integrated AI editor rather than a terminal grid manager. ## Overview Paneherd is a VS Code and Cursor extension that puts up to nine AI coding agent CLIs, including Claude Code, OpenAI Codex, GitHub Copilot CLI, Cursor CLI, Gemini CLI, OpenCode, Qwen Code, Antigravity, and Factory Droid, into a single resizable grid inside the editor. The core idea is simple: instead of switching between terminals to find which agent is waiting on you, one sidebar shows every pane's live status, collects Claude Code permission prompts in a shared inbox, and lets you answer Allow, Always, or Deny without leaving your current file. What separates Paneherd from just opening multiple terminals is the coordination layer on top. A conflict guard intercepts edits when two agents are about to touch the same file. A loop alarm fires when an agent appears stuck. A task queue holds prompts and hands them to the first agent that finishes its current task. Version 0.17 added a merge queue that only lands a finished worktree after its tests pass, plus a cross-vendor review where the other agent reviews the branch before merge. The phone remote is a practical differentiator: scan a QR code, click Allow in VS Code, and your phone's browser shows every pane's status and lets you send prompts or answer permission requests from anywhere. The connection is end-to-end encrypted; the relay passes sealed messages and cannot read them. No app to install. Paneherd does not proxy any models and does not touch your API keys. It runs the agents' own CLIs locally on your machine, on the subscriptions you already hold. That architecture means it adds no per-token cost and introduces no new trust surface between you and the model providers. ## Who should use it Paneherd is built for developers who already run Claude Code alongside at least one other agent CLI and find themselves constantly switching terminals to check which session needs attention. A solo developer running Claude Code for architecture work and Codex for a second opinion on the same codebase gets immediate value from the shared inbox and conflict guard. Teams using git worktrees per feature branch will find the merge queue in 0.17 particularly useful, since it holds a branch until tests pass before landing it. If you use a single agent and want deep editor integration, autocomplete, and inline suggestions rather than a terminal grid, look at Cursor instead. Cursor is an AI-first editor that understands your whole codebase and handles the full editing loop natively. If you want an autonomous agent that takes a GitHub Issue all the way to a tagged release without manual terminal management, Orbi is the more appropriate tool. ## FAQ Q: Does Paneherd require me to create an account? A: No. The free tier needs no account at all. The Pro license key is purchased at paneherd.com and activated directly in VS Code or Cursor; the license check runs offline with no network call required after activation. Q: How much does Pro cost and is there a trial? A: Pro is a one-time $2.99 purchase with a perpetual license key. There is a 7-day free trial you can start from the sidebar without entering payment details, and a 14-day refund window after purchase. Q: Which editors does Paneherd support? A: Paneherd runs in VS Code 1.95 or later and in Cursor and other editors built on the VS Code platform. It is not available for JetBrains IDEs, Neovim, or any editor outside the VS Code family. Q: Does Paneherd proxy my API calls or see my model credentials? A: No. Paneherd runs each agent's own CLI locally and communicates with the extension over localhost. It does not proxy model calls, does not handle API keys, and has no telemetry. Q: Which agents does Paneherd support? A: As of version 0.17, Paneherd supports Claude Code, OpenAI Codex, GitHub Copilot CLI, Cursor CLI, Gemini CLI, OpenCode, Qwen Code, Antigravity, and Factory Droid. The permission inbox and conflict guard are Claude Code-specific features; other agents get status display and prompt broadcasting. Q: What does the phone remote actually let me do for free versus Pro? A: On the free tier, the phone remote shows every pane's status and lets you send prompts to selected panes. Pro adds the ability to answer Claude Code permission prompts (Allow, Always, or Deny) from the phone and receive a buzz when a pane needs attention.