Website review
This review may include AI-generated content and is not individually verified. Domain, certificate and other facts come from the observed data on this page.
What is WTF?
WTF (wtfutil) is a personal information dashboard that runs in your terminal. Rather than switching between browser tabs and web apps, you configure a single grid of modules and view live status from each inside one terminal window. Per WTF, the project describes itself as "the personal information dashboard for your terminal," and it ships a large catalog of modules covering developer tools, cloud services, monitoring, and everyday data.
What it is good for
Each module pulls from one source and renders it as a small tile. The catalog on the site spans several practical groups:
- Code and CI: GitHub, GitLab, Jenkins, CircleCI, TravisCI, BuildKite, Gerrit, Mercurial.
- Work tracking and incidents: Jira, Trello, Asana, Todoist, PagerDuty, OpsGenie, VictorOps, Zendesk.
- Monitoring and security: Datadog, New Relic, UptimeRobot, Healthchecks, Rollbar, Have I Been Pwned.
- Cloud and infrastructure: Kubernetes, Docker, DigitalOcean, Azure DevOps, Pihole, IP address and IPInfo lookups.
- Everyday information: Weather, clocks, stocks and crypto, HackerNews, subreddits, RSS feeds, Spotify, and a system resource-usage tile.
You arrange these in a grid layout defined in configuration files, so the dashboard is whatever mix of tiles you choose. The underlying project is open source.
Who it suits, and the trade-off
WTF fits people who already live in a terminal—engineers, SREs, and ops staff—and want a glanceable overview of builds, tickets, alerts, and metrics without leaving the shell. The trade-off is that it is not a point-and-click product: you edit config files to add modules and set layouts, and you supply credentials for each service you connect. If you prefer a visual dashboard you can drag around in a browser, this will feel like extra setup.
Next step: open the Getting Started and Modules sections of WTF and list the five or six sources you check most often. If most of them appear in the module catalog, start with those tiles in a simple grid; if your key tools are missing, a browser-based dashboard will likely serve you better.
How do I install and set up WTF in my terminal?
WTF is a personal dashboard that runs inside your terminal, showing information from many services as a grid of modules. Installation and setup follow the same two steps regardless of platform: install the binary, then generate and edit a configuration file.
Install
The project's documentation lists installation instructions under Getting Started, including a Quick Start and a separate Uninstallation page. In practice, most people install the prebuilt binary for their operating system and put it on their PATH; the docs also describe alternative install methods. After installing, run the tool once — it will look for a config file and, if none exists, help you create one.
Configure
Configuration is the main work, and it happens in a YAML config file. The docs organize it into four parts:
- Configuration Files — where the config lives and how it is loaded.
- Global Settings — refresh intervals and other app-wide options.
- Common Settings — options shared by many modules.
- Grid Layout — how modules are arranged into rows and columns.
A practical first config is deliberately small: a clock, the weather, and one service you already use, such as GitHub or HackerNews. Get the grid right with those three, then add modules one at a time so a broken module is easy to isolate.
Choose modules
The module list is broad — developer tools (GitHub, GitLab, Jenkins, CircleCI, Jira, Kubernetes), infrastructure and monitoring (Datadog, PagerDuty, UptimeRobot, Resource Usage, Docker), and personal feeds (Weather, Stocks, Todoist, Spotify, Twitter). Each module has its own settings and usually needs an API token or credentials.
Two trade-offs are worth deciding up front:
| Approach | Good for | Cost |
|---|---|---|
| Few modules, large cells | A quick morning check of 3–5 things | Less at-a-glance coverage |
| Many small modules | Ops and on-call users watching several systems | Denser layout, more tokens to manage, slower refresh |
A concrete next step
Pick one service you check daily — say GitHub notifications — find its module page in the docs, copy the example config block, and add your token. Run WTF, confirm the module renders, then repeat. If you want a second reference for terminal dashboards generally, see GitHub for related projects, but the WTF documentation itself is the authoritative source for module settings and layout syntax.
Which WTF module should I use to monitor my GitHub or GitLab repositories?
For GitHub or GitLab repository monitoring, WTF provides separate GitHub and GitLab modules rather than a single combined one. Choose the module that matches where the repositories live; if you use both services, add both modules to your dashboard.
What each module is for
| Your situation | Module to use | Typical things to surface |
|---|---|---|
| Repositories, issues and pull requests are on GitHub | GitHub | Open PRs, assigned issues, review requests, repository activity |
| Work is hosted on GitLab (including self-managed) | GitLab | Merge requests, assigned issues, project activity |
| You contribute to both | GitHub and GitLab side by side | A combined view across both accounts |
The practical distinction is workflow, not features: the GitHub module suits teams whose review process runs through pull requests and GitHub notifications, while the GitLab module fits organisations using merge requests and GitLab projects. If you only occasionally check the other service, keeping one module avoids a cluttered grid.
A concrete setup
Suppose you maintain three GitHub repositories and one GitLab project for a client. Add the GitHub module configured for your username and the repositories you care about, and the GitLab module pointing at the client's project. Place them in adjacent grid cells so pull requests and merge requests sit together at a glance. The page lists configuration and grid layout sections for details on how modules are placed and configured.
Decision criterion
Start with the service that generates most of your review requests. Add the second module only when you find yourself switching to a browser for it. For broader context alongside code activity, the GitHub and GitLab modules pair naturally with issue trackers such as Jira if that is where your tickets live.
Next step: open the module's page in the WTF documentation (under Modules) to see its configuration options before adding it to your config file.
How do I configure WTF modules and the grid layout in the config file?
WTF is configured through a YAML config file, and the two things you'll edit most are the grid layout and the module definitions. The config file is typically at ~/.config/wtf/config.yml (WTF creates a default one on first run). The page's own documentation splits this into "Configuration Files," "Global Settings," "Common Settings," and "Grid Layout," with each module having its own settings page under "Modules."
Grid layout
The layout is a set of columns, each containing rows, each row holding one or more modules. The module names in the grid must match the names you define in the mods section.
wtf:
colors:
border:
focusable: darkslateblue
focused: orange
normal: gray
grid:
columns:
- width: 2
rows:
- rate: 1
widgets: [clock, weather]
- rate: 1
widgets: [bargraph]
- width: 1
rows:
- rate: 1
widgets: [github]
- rate: 1
widgets: [todo]
width sets relative column width and rate sets relative row height, so you can give a busy module more space without absolute pixel values. Add a focusable: true to a widget if you want keyboard focus for scrolling or interaction.
Module definitions
Each module gets an entry under mods, keyed by the name you used in the grid, with a type and its own options:
mods:
clock:
type: clocks
colors:
rows:
even: "lightblue"
odd: "white"
github:
type: github
repositories:
- wtfutil/wtf
refreshInterval: 5m
todo:
type: todo
filename: ~/.config/wtf/todo.yml
The type selects the module implementation (the docs list types such as clocks, github, todo, weather, bargraph, cmdrunner, stocks, and many more); everything else is module-specific. refreshInterval and color settings are common across most modules.
Practical guidance
- Start from the generated default config and change one column or module at a time, restarting WTF to check the result. A typo in a widget name in the grid is the most common cause of a blank or missing module.
- Keep secrets out of the main file where a module supports it — for example, point token-based modules at an environment variable or a separate file rather than pasting keys into
config.yml. - Use
cmdrunnerfor anything not covered by a built-in module; it runs a shell command and displays the output, which is often simpler than writing a custom module. - If you want to share or version the config, keep it in a dotfiles repo and symlink it into place.
For the full list of module types and their exact option names, the project's documentation at WTF is the authoritative reference, and the source repository is linked from there.
Can I build a custom WTF module for a service that isn't supported?
Yes. WTF is designed to be extended with your own modules, so a service that has no built-in module can still appear on your terminal dashboard. The page lists a "Creating New Modules" section under Development, which is the documented path for adding functionality beyond the bundled modules.
What a custom module is for
Build one when you want a specific service's data on the same grid as your clocks, resource usage, Git activity and other widgets. Typical cases:
- An internal API or homegrown tool with no public module.
- A niche SaaS product the project doesn't cover.
- A filtered or reshaped view of data you already get elsewhere.
How to approach it
- Start from the "Creating New Modules" documentation rather than improvising, since it defines the module interface and configuration conventions.
- Model the module on the closest existing one. The module list is broad — covering categories like cloud, CI, monitoring, source control, feeds, weather and finance — so a similar widget usually exists to copy.
- Decide what the module displays: a single summary line, a short list, or a small table. Terminal space is the real constraint, so keep output compact.
- Reuse the global and common settings and the grid layout system so your module behaves like the rest.
Trade-offs to weigh
| Option | Effort | Control | Maintenance |
|---|---|---|---|
| Custom module | Higher, requires coding | Full control over data and display | You own updates |
| Existing general-purpose module (e.g. command runner or feed reader) | Low | Limited to what that module exposes | Handled upstream |
| External script plus a text/file-based module | Low to moderate | Moderate | Mostly yours |
If your need is "run a command and show its output," a command-runner or text-file module may be enough and avoids writing code. A true custom module pays off when you need structured data, periodic refresh and formatting that generic modules can't produce.
For background on the project and its module conventions, see WTF - the terminal dashboard.
How does WTF compare to other terminal dashboards like sampler or glances?
WTF (wtfutil.com) is a terminal dashboard built around a configurable grid of modules rather than a single fixed view. Its page lists dozens of integrations—GitHub, GitLab, Jira, Kubernetes, Docker, cloud services, weather, stocks, feeds and more—so its main appeal is pulling work and service status into one terminal screen. Sampler and Glances take different approaches: Sampler focuses on shell-command-driven visualizations and charts, while Glances is primarily a system and resource monitor with an optional web view. WTF is the most integration-oriented of the three; Glances is the most out-of-the-box for host monitoring; Sampler is the most flexible for custom command output.
How to choose
- Monitoring servers and local resources: Glances is the natural starting point, since resource metrics are its core purpose.
- Watching CI, tickets, repos and cloud services: WTF's module catalog fits this better than a resource monitor.
- Building charts from arbitrary commands: Sampler suits ad-hoc visualizations of command output.
- A single terminal "home screen" for mixed signals: WTF's grid layout is designed for this, at the cost of configuring each module.
A practical test: list the five things you check each morning. If most are services and repositories, try WTF; if most are CPU, memory and disk, start with Glances. You can also run more than one—for example, Glances for the host and WTF for project and service status—though screen space and configuration effort are the trade-offs. For the official projects, see WTF and Glances.
User reviews (0)