RunHQ is a local-first developer cockpit. It auto-discovers your repos, runs services and stacks, opens log tabs, watches ports, scans CVEs, tracks git status, and lets AI triage your workspace. Native desktop app for macOS, Linux and Windows, MIT-licensed, written in Rust and Tauri.

How is RunHQ different from Foreman, Overmind, mprocs or Concurrently?

Procfile runners like Foreman, Overmind and mprocs cover one Procfile per repo. RunHQ covers the whole machine: services across every repo, port watchdog, CVE scanner, git status, AI triage, and a native cross-platform GUI.

Does RunHQ replace Docker Desktop?

For local development, RunHQ often replaces Docker Desktop. It runs services natively (Node, Go, Rust, Python and more), but can also start Docker Compose stacks. Teams typically keep Docker for production-parity tests and use RunHQ for the daily inner loop.

Does my code or AI prompt ever leave my machine?

No. RunHQ has zero telemetry, no account, and no built-in cloud sync. AI triage uses any OpenAI-compatible endpoint you provide, including local models like Ollama, LM Studio or llama.cpp. All data persists to a local SQLite file under ~/.runhq.

Which runtimes and package managers does RunHQ auto-discover?

RunHQ auto-discovers ten runtime families: Node.js, Bun, Deno, Go, Rust, .NET, Python (Poetry, uv, pip), Java (Maven and Gradle), Ruby, PHP, and Docker Compose. It reads package.json, go.mod, Cargo.toml, pyproject.toml, compose.yaml and similar manifests to infer start commands.

How does RunHQ work alongside VSCode, Cursor or JetBrains IDEs?

RunHQ complements editors like VSCode, Cursor, Zed, Sublime, Xcode and JetBrains. It runs and observes the services your editor is editing, with one-click "open in IDE" buttons for every workspace. RunHQ does not replace editor tasks or debug configurations.

Is RunHQ free? What is the licence?

Yes. RunHQ is free and open source under the MIT licence. There is no paid tier, no seat pricing, and no account requirement. The source is on GitHub at github.com/erdembas/runhq and every release is signed.

How heavy is RunHQ on system resources?

RunHQ targets under 100 MB of RAM at idle. It ships as a Tauri and Rust native binary, with a React renderer over Rust IPC. Cold start to first paint is under 300 milliseconds on Apple Silicon.

v0.10.0local-first dev cockpit

One cockpit.
for every service, every log line, every stuck port, every CVE — for your whole machine.

Your editor edits code. RunHQ runs it.

Install free Watch the 5-min demo

Native · Tauri + Rust
< 100 MB RAM
10 runtimes
MIT
Zero telemetry

The full cockpit

One window. Every project. Every running command.

What you’re looking at below is the actual React surface RunHQ ships — sidebar, tab strip, dashboard body and status bar pulled straight from the desktop package and re-rendered here with a fixture workspace.

RunHQ

Search services, stacks, actions…⌘K

3 running

Workspace·7 services·v0.10.0·Last scan 1h ago

1service need attention

287 MB 0.0% 4 ports

Running hot3 running

0.0%287 MB

acme-web
178 MB0.0%
acme-compose
80 MB0.0%
acme-api
29 MB0.0%

Search projects/

Acme Projects4● 3 on

acme-apiGO

All clear

1h ago

2 commands·run, main

main

0.0%29 MB

[2026-05-05T13:18:24.860] background scan complete

[2026-05-05T13:18:18.740] background scan complete

[2026-05-05T13:18:14.560] background scan complete

acme-composeDOCKER

All clear

1h ago

2 commands·compose-up, compose-down

main

0.0%80 MB

[1 t e n a c e - p o s t g r e s | 1] db migrations · idle

[1 t e n a c e - p o s t g r e s | 1] checkpoint · 217 frames

[1 t e n a c e - p o s t g r e s | 1] autovacuum · 0 dead rows

acme-webNODE

All clear

2 commands·dev, lint

main

0.0%178 MB

[2026-05-05T13:18:17.792] [vite] page reload (build)

[2026-05-05T13:18:14.382] [vite] hmr update src/...

[2026-05-05T13:18:13.301] [vite] dev server ready

billing-workerPYTHON

Idle · no dependencies missing

1h ago

1 commands·main

main

Tooling1

internal-cliRUST

Idle · no dependencies missing

1h ago

1 commands·run

main

Unassigned2

design-systemNODE

Idle · no dependencies missing

1h ago

1 commands·storybook

main

docs-siteNODE

Idle · no dependencies missing

1h ago

1 commands·dev

main

3 running

4 idle

0.0%287 MB4 ports· Ask AIAI Settings Darkv0.10.0 · Release notes

Watch the loop · Full demo

The full daily loop, in five minutes.

Discover repos, start the stack, chase logs, triage CVEs, ask AI for next actions — captured straight off a real workspace. Click play and you’re in.

Auto-discovers

Node.jsBunDenoGoRust.NETPythonJava · MavenJava · GradleRubyPHPDocker Compose

Product loop · Discover · Run · Triage

The daily loop RunHQ compresses.

Find the repo, remember how it boots, start the stack, chase logs, check ports, then decide what deserves attention. RunHQ turns that routine into one persistent local cockpit.

One scan, no memory tax.

RunHQ walks your dev folders, infers runtimes and start commands from manifest files, and keeps the registry local. Open the app and your projects are already a workspace, not a directory hunt.

10 runtime families · Node, Bun, Deno, Go, Rust, .NET, Python, Java, Ruby, PHP, Docker
Manifest-aware commands · package.json, go.mod, Cargo.toml, pyproject.toml, compose.yaml
Persistent local registry · ~/.runhq/config.json plus SQLite snapshots

RunHQ · ~/code · scanning…

Walking ~/code

7 projects indexed·cached to ~/.runhq/config.json

3 node1 go1 rust1 .net1 docker

00:00acme-webnodepnpm dev
00:01acme-apigogo run ./cmd/api
00:02internal-clirustcargo run
00:03finsel-billingdotnetdotnet run
00:04data-jobspythonuv run train.py
00:05design-systemnodepnpm storybook
00:06acme-stackdockerdocker compose up

Why · The local-first dev cockpit

Your editor edits code. RunHQ coordinates the environment.

01 · The status quo

Terminal tabs are hidden state.

Foreman runs one Procfile. Your IDE remembers recents. Docker Desktop shows containers. The rest is a pile of terminal tabs, shell history and memory.

acme-web :3000NODE

acme-api :8080GO

docker composeDOCKER

data-jobs trainPYTHON

4 terminals, 3 runtimes, 1 broken port, zero shared state.

02 · RunHQ

RunHQ gives local dev one surface.

Projects, services, logs, ports, git state, CVEs and stale scans live together. The cockpit answers “what is happening on this machine?”

acme-platformDOCKER main

Running hot3 running

0.0%287 MB

acme-web
178 MB0.0%
acme-compose
80 MB0.0%
acme-api
29 MB0.0%

3 on 4 idle0.0%·287 MB

03 · The AI difference

AI gets the missing context.

File-level AI can write code. RunHQ can explain the workspace around that code: which project is stale, which command failed, which CVE should go first.

1 cve 3 stale 2 dirtyworkspace pulse

> which projects need me this week?3 repos haven’t moved in 90d. acme-api has a critical CVE in axios. internal-cli is 12 packages outdated.

Features · Run · Observe · Triage · Think

Everything outside the editor, finally in one place.

RunHQ is the local dev cockpit: launch the stack, watch the logs, clear port conflicts, read workspace health and ask AI for next actions without leaving the same surface.

Run

acme-apiRUST

All clear

1 commands·dev

main

4.2%29 MB

Start services, stacks and custom commands without reopening the same terminal tabs every morning.

SIGTERM → grace → SIGKILL
Multi-command per service
Port watchdog
One-click kill port
Embedded xterm.js
Open in 8 IDEs
⌘K palette
⌘⇧K global hotkey

Graceful shutdown, multi-command services, one-click IDE launch.

Observe

›_ api · go run ./cmd/api

RUN

14:02GET /healthz · 200

14:02!WARN n+1 query · orders.handler

14:03POST /orders · 201

Command logs open as tabs, render like a real terminal, and stay searchable while services run.

Ring buffer · 10k/svc
Virtualized list
ANSI color · live filter
Activity timeline
Daily / weekly rollups
Standup export
Right-click → AI triage
SQLite persistence

Search, follow, copy lines, and send suspicious output to AI.

Triage

1 cve 3 stale 2 dirty

12%·287 MB

↑ axios CVE blocks deploy12 outdated

See stale scans, outdated packages, CVEs and dirty git across the whole local portfolio before they become surprises.

npm · cargo · pip audit
Persistent scan cache
Per-card freshness
Per-project rescan
Scan delta badges
"N stale" pill
Worst-offenders panel
Triage drawer

Cold start paints chips in < 100 ms from SQLite.

Think

> what should I fix first?acme-api · axios CVE blocks deploy. Patch first.

local · qwen2.5-coder128 tok

Ask questions against the whole workspace with your own OpenAI-compatible endpoint, local or cloud.

Project · Why?
Workspace report
Per-CVE deep-dive
In-place commit gen
Diff explain
Log triage
5-tab chat · SQLite
tiktoken-rs token meter

Point it at a local model — code, diffs, logs never leave the machine.

Alternatives · Comparison

How RunHQ compares to Foreman, Overmind, mprocs and Concurrently.

Procfile runners cover one process file. Editor task runners cover one workspace. RunHQ covers the whole machine — every service, every port, every repo’s git and CVE state, in one local-first cockpit.

RunHQ · process-manager comparison

6 tools · 8 capabilities

Capability matrix comparing RunHQ to other developer process managers and local service runners. Columns are capabilities, rows are tools.
Tool	Multi-processRun many services at once	Multi-runtimeNode, Go, Rust, Python, …	Searchable logsANSI, ring buffer, follow	Port watchdogSee & kill stuck ports	CVE / outdatedPer-project audit	Git statusAcross all repos	AI triageBYO endpoint, local-first	Native GUImacOS · Linux · Windows
RunHQLocal dev cockpit · Tauri + RustMIT · < 100 MB RAM · zero telemetry	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
ForemanProcfile runner · Ruby CLI	Yes	Partial	Partial	No	No	No	No	No
Overmind / Hivemindtmux-backed Procfile runner	Yes	Partial	Yes	No	No	No	No	No
mprocsTUI process manager · Rust	Yes	Partial	Yes	No	No	No	No	No
Concurrently / npm-run-allnpm script multiplexer	Yes	No	Partial	No	No	No	No	No
VSCode tasks / launch.jsonEditor task runner	Partial	Yes	Partial	Partial	No	Partial	Partial	No

Capability data sourced from each project’s public README / docs as of May 2026. Open a PR if a row is out of date — github.com/erdembas/runhq.

Install · macOS · Linux · Windows · v0.10.0

Install once. Stop rebuilding your setup.

macOS

11+

.dmgIntel Apple Silicon

Cask postflight clears Gatekeeper quarantine. First launch is one click.

Linux

x64 & ARM64

.debx64 ARM64

.rpmx64 ARM64

AppImagex64 ARM64

Every artifact is signed with the same minisign key the in-app updater verifies.

Windows

10 / 11

.exex64 ARM64

.msix64 ARM64

The in-app updater keeps it current — no reinstall needed between releases.

Need an older version, the .sig alongside a binary, or the Tauri updater manifest? Browse every release on GitHub. Prefer source? pnpm tauri:dev — Rust + pnpm, that’s the full toolchain.

FAQ · Developer questions

Questions developers ask before they install.

Eight straight answers about runtimes, privacy, licence, IDE adjacency and how RunHQ sits next to the tools you already use.

01What is RunHQ?
Answer
RunHQ is a local-first developer cockpit for your laptop. It auto-discovers the repos in your dev folder, lets you start whole stacks with one click, opens a log tab per command, watches ports, scans dependencies for CVEs, and lets AI triage the workspace. Native desktop app, MIT-licensed, written in Rust + Tauri.
02How is RunHQ different from Foreman, Overmind, mprocs or Concurrently?
Answer
Procfile-style runners (Foreman, Overmind, Hivemind) and TUI process managers (mprocs) cover one Procfile per repo and stop there. RunHQ covers the whole machine — services across every repo, port watchdog, CVE + outdated-dependency scanner, git status, AI triage, and a real native GUI. See the comparison matrix for the cell-by-cell breakdown.
03Does RunHQ replace Docker Desktop?
Answer
For local development, often yes. RunHQ runs your services natively (Node, Go, Rust, Python, …) without containerising them, but it can also start Docker Compose stacks the same way it starts a `pnpm dev`. Many teams keep Docker for production parity tests and use RunHQ for the day-to-day inner loop.
04Does my code or AI prompt ever leave my machine?
Answer
Never by default. RunHQ has zero telemetry, no account, and no built-in cloud sync. AI triage uses any OpenAI-compatible endpoint you point it at — that can be a local model (Ollama, LM Studio, qwen2.5-coder, llama.cpp) or your own cloud key. Logs, scans, and chat tabs persist to a local SQLite file under ~/.runhq/.
05Which runtimes and package managers does RunHQ auto-discover?
Answer
Ten runtime families today: Node.js, Bun, Deno, Go, Rust, .NET, Python (Poetry / uv / pip), Java (Maven + Gradle), Ruby, PHP, plus Docker Compose. RunHQ reads package.json, go.mod,Cargo.toml, pyproject.toml, compose.yaml and friends to infer start commands.
06How does RunHQ work alongside VSCode, Cursor or JetBrains IDEs?
Answer
Your editor edits code; RunHQ runs it. The cockpit watches the same workspace your editor has open and exposes one-click “open in IDE” buttons for VSCode, Cursor, Zed, Sublime, Xcode, JetBrains and any terminal — picking up where you left off without re-typing paths. RunHQ does not replace your editor or its run/debug configurations.
07Is RunHQ free? What is the licence?
Answer
Yes, free and open source under the MIT licence. No paid tier today, no seat-based pricing, no account walls. The full source lives on github.com/erdembas/runhq and every release is signed with the same minisign key the in-app updater verifies.
08How heavy is RunHQ on system resources?
Answer
RunHQ targets under 100 MB of RAM idle and ships as a Tauri + Rust native binary — the same engine that powers tools like Linear’s desktop app and Spacedrive. The renderer is React, but every cross-process call is a Rust IPC, not Electron-style Node-bridged JS. Cold start to first paint is sub-300 ms on an M-series Mac.

Got a question that isn’t here? Open a discussion on GitHub.

Stop carrying your dev setup in terminal tabs. Start from RunHQ.

Free · MIT · macOS · Linux · Windows · < 100 MB RAM

Get RunHQ Star on GitHub

No account. No credit card. No telemetry.

One cockpit.for every service.for every service, every log line, every stuck port, every CVE — for your whole machine.

One window. Every project. Every running command.

1service need attention

The full daily loop, in five minutes.

The daily loop RunHQ compresses.

One scan, no memory tax.

Your editor edits code. RunHQ coordinates the environment.

Terminal tabs are hidden state.

RunHQ gives local dev one surface.

AI gets the missing context.

Everything outside the editor, finally in one place.

Run

Observe

Triage

Think

How RunHQ compares to Foreman, Overmind, mprocs and Concurrently.

Install once. Stop rebuilding your setup.

Questions developers ask before they install.

Stop carrying your dev setup in terminal tabs. Start from RunHQ.

One cockpit.
for every service, every log line, every stuck port, every CVE — for your whole machine.