feed · watching · stats · help with your tokens
almogdepaz/wolfpack
Wolfpack is a self-hosted browser terminal manager for controlling AI coding agents remotely over Tailscale.
Wolfpack solves the problem of managing remote AI coding agents by providing a self-hosted control room that allows users to monitor and control persistent terminal sessions from a desktop or phone without being physically present at the host machine. Wolfpack uses a Rust PTY broker to keep terminal sessions persistent independently of the web server, ensuring sessions survive browser closures or web server restarts.
Similar
- jyotipmahes/ML-Interview-Prep
A repository containing Machine Learning Engineer interview preparation materials and from-scratch algorithm implementations.
- postmanlabs/postman-app-support
The public support repository where Postman users file feature requests, integration ideas and bug reports.
- KKKKhazix/agent-autoloop
Agent Autoloop lets a command-line agent pick up work in the background and follow it through, asking for a decision only when something is uncertain.
- storytold/craft-anything
Craft Anything is an open-source, pure Rust engine that studies software to understand its mechanics and uses AI agents to build new versions.
- lookfree/cc-harness
A desktop workbench for Claude Code that visualizes subagent topology, token costs, and configuration to improve observability.
- caru-ini/stack-diagram
An agent skill that generates architecture diagrams with official logos from a one-sentence description of a tech stack.
- soml-lang/soml
SOML is a configuration format for humans that keeps JSON's data model and strictness while dropping braces, quotes and commas.
- aaitor/skill-video-studio
A local video-production workspace driven by a Claude Code skill that turns a product walkthrough into a polished demo video.
sources
Claims below include README evidence. A matching quote or worker agreement does not establish that a claim is correct. Category majority among 2 workers. Version 1.
Wolfpack is a self-hosted browser terminal manager for controlling AI coding agents remotely over Tailscale.
Wolfpack is a self-hosted control room built for remote and mobile use
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Wolfpack solves the problem of managing remote AI coding agents by providing a self-hosted control room that allows users to monitor and control persistent terminal sessions from a desktop or phone without being physically present at the host machine.
Run coding agents remotely—from your desktop or phone. Keep their persistent terminals on your own machines, then monitor and control them even when you are away from your desk
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Wolfpack uses a Rust PTY broker to keep terminal sessions persistent independently of the web server, ensuring sessions survive browser closures or web server restarts.
Sessions live in a Rust PTY broker, not the web server, so closing the browser or restarting only the web server does not end them
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Supports built-in coding agents including Claude Code, Codex, Gemini CLI, Cursor, and Pi, with shell as a fallback.
Built-in coding-agent provider choices: Claude Code, Codex, Gemini CLI, Cursor, Pi
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Provides a mobile PWA for monitoring agent status, opening live terminals, and responding to prompts remotely.
From the phone PWA, you can see which agents need attention, open a live terminal, and respond without returning to the host machine
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Operates over a private Tailscale network without requiring a hosted relay or external account.
Connect directly over your private Tailscale network without a Wolfpack-hosted relay or account
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Enables multi-machine session control across trusted peers on the Tailnet.
trusted multi-machine session control
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Displays needs-input states for agents and supports optional phone notifications.
needs-input states and optional phone notifications
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Allows configuration of custom commands or wrappers on the PATH for agent integration.
custom commands or wrappers on `PATH` are supported separately
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Includes agent skills for controlling sessions via compatible agent harnesses.
wolfpack-tailnet-control works with all agent harnesses that support Agent Skills
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Supports CLI automation for spawning child agents with specific harnesses and plans.
wolfpack agent spawn project-name --harness claude --plan .plans/000-review.md --notify-parent --json
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Enables remote monitoring and control of AI coding agent terminals from desktop or mobile devices.
Run coding agents remotely—from your desktop or phone. Keep their persistent terminals on your own machines, then monitor and control them even when you are away from your desk.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Supports multiple AI coding agents including Claude Code, Codex, Gemini CLI, Cursor, and Pi, with shell fallback.
Built-in coding-agent provider choices: Claude Code, Codex, Gemini CLI, Cursor, Pi. Shell is the always-available fallback, and custom commands or wrappers on PATH are supported separately.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Connects directly over a private Tailscale network without requiring a hosted relay or account.
Connect directly over your private Tailscale network without a Wolfpack-hosted relay or account.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Provides a phone PWA interface to view agent status, open live terminals, and respond remotely.
From the phone PWA, you can see which agents need attention, open a live terminal, and respond without returning to the host machine.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Allows users to monitor sessions, identify agents waiting for input, and take over terminals remotely.
Monitor your sessions, spot agents waiting for input, and take over a live terminal from the phone PWA.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Displays local sessions and verified peers on a multi-machine dashboard and desktop sidebar.
The multi-machine dashboard and desktop sidebar show only local sessions and currently ready, handshake-verified Wolfpack peers.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Integrates with agent harnesses supporting Agent Skills via the wolfpack-tailnet-control skill.
wolfpack-tailnet-control works with all agent harnesses that support Agent Skills.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Allows creation of top-level project sessions with initial instructions via CLI commands.
wolfpack session create project-name --harness pi --plan .plans/000-task.md --json
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Provides shell-level control over configured projects within a trusted Tailnet environment.
wolfpack gives browser users shell-level control over configured projects. Keep it private to a trusted Tailnet.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Ensures sessions remain attached to a running broker even during service restarts.
Sessions on a running broker remain attached. See what the installer does for the full lifecycle boundary.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Fails closed for invalid or unreachable targets without falling back to localhost connections.
Invalid, incompatible, redirected, timed-out, or unreachable targets fail closed without localhost fallback.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Requires a trusted Tailnet environment because it grants shell-level control to browser users.
wolfpack gives browser users shell-level control over configured projects. Keep it private to a trusted Tailnet
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Sessions may be terminated by a broker restart or login-service reinstallation.
a broker restart or login-service reinstallation can terminate them
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Lacks an inter-session authorization layer, relying on global API auth policies.
Session control follows the ordinary global API auth policy when configured and has no inter-session authorization layer
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Lacks an inter-session authorization layer and relies on global API auth policy for session control.
Session control follows the ordinary global API auth policy when configured and has no inter-session authorization layer; read the full trust model.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Platform binaries do not expose skills as files, requiring manual symlink or copy instructions.
Platform binaries do not expose skills as files; detailed safe symlink and copy instructions are in agent skills.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Only accepts existing absolute directories for project selection, rejecting relative paths on the server.
Relative CLI paths are resolved locally; the server accepts and canonicalizes only existing absolute directories.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Excludes generic, offline, malformed, or unreachable Tailnet candidates from the control room view.
Generic, offline, malformed, and unreachable Tailnet candidates stay out of the control room; Settings → Machines retains bounded discovery diagnostics.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Requires configuring device/user ACLs and considering JWT if other people share the Tailnet.
If other people share the Tailnet, configure device/user ACLs and consider JWT.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
Limitation: Requires reporting suspected vulnerabilities through private security channels, not public issues.
Report suspected vulnerabilities through private security reporting, never a public issue or discussion.
Quote matched the README or GitHub description at extraction. Same claim and quote from 1 worker(s). Interpretation not reviewed by a maintainer.
remote-access · developer-tools