The terminal that holds your whole workday

Features

• Real shells. Each pane runs its own shell process, so a long build in one never blocks another.
• Unlimited splits right or down, resized by dragging the divider.
• Per-pane settings - font, size, colours and title - or one global look for all.
• Open a folder straight into a new pane with ⌘O.
• Safe closing. Closing the window asks first, so a stray click never ends running work.
• All shell keys pass through - Ctrl C, Ctrl R, Ctrl Z reach the shell untouched.

How to use

1. Click + New terminal (or ⌘⇧T) to open a shell.
2. Click Split right or Split down to divide the active pane.
3. Drag the line between panes to resize them.
4. Jump between panes with ⌘1…⌘9, or ⌘⇧[ / ⌘⇧].
5. Close the active pane with ⌘W; toggle the sidebar with ⌘B.

Features

• One tab per open item - code projects, remote servers, browsers, file windows, and tools such as the API client, Docker, CI/CD or the Vault.
• Nothing is lost when you switch. A project put aside keeps its open files, unsaved edits, expanded folders and scroll position.
• Home puts every panel aside at once and shows your terminals, which keep running underneath.
• No duplicates: opening a project that is already open simply brings its tab forward.
• Closing asks first if there is unsaved work.
• The bar appears only when something is open, so terminals keep the full height otherwise.

How to use

1. Open a project from Code, or click + Project in the bottom bar to add another one beside it.
2. Click a tab to bring that project or tool to the front.
3. Double-click the tab in front to put it aside again.
4. Click Home to go back to your terminals.
5. Click the × on a tab to close it for good.

Tip: hover over a project tab to see its full folder path - useful when two projects share the same name.

Features

• Favourites, places and drives in the sidebar, one click to open.
• List and grid views with sortable, resizable columns - name, dates, size, kind.
• Copy, move, rename, delete with the usual shortcuts and a right-click menu.
• Zip and unzip, including password-protected archives.
• Search the current folder and everything below it.
• Quick preview of images, text and documents without opening another app.

How to use

1. Click Files in the toolbar.
2. Choose a place on the left; double-click a folder to open it.
3. Click a column header to sort; drag a header edge to resize it.
4. Select items, then right-click for Compress, Copy, Move or Rename.
5. Select a zip and choose Unzip… - enter the password if it has one.

Features

• 102 commands, each with a description, a worked example and related commands.
• Run (▶) sends a complete command to the active pane and executes it.
• Insert (⤓) types a command that needs a value - such as <branch> - and leaves it for you to finish.
• Editable command box: change the command before you run it.
• System monitor: CPU, memory and temperature graphs, plus load, disk, processes and uptime.

How to use

1. Press ⌘⇧P or click >_.
2. Type to search, for example “port” or “docker logs”.
3. Select a command to see what it does.
4. Edit the command box if needed, then click Run or press ⌘↩.

Safe by design: commands that need sudo or a placeholder are never run directly - they are typed for you to confirm.

Features

• Four assistants - Claude Code, Codex, Kiro and Antigravity - each in its own tab.
• Session history by project, with a clean, searchable transcript.
• Resume or start a session in a split pane; when the AI exits, the pane is still a normal shell.
• Open the project in the built-in code editor or in VS Code, Cursor or VSCodium.
• Edit Claude Code's settings safely - invalid JSON is refused and a backup is kept.
• Several Claude accounts can be saved, so you can switch between them - for example, your personal and work accounts.

How to use

1. Click AI and choose an assistant, or press ⌘⇧C for Claude Code.
2. Pick a project on the left, then a session.
3. Read the transcript; press ⌘F to search it.
4. Click Resume in split terminal - or New session here to start fresh.
5. To add a folder that has no history yet, click New… above the project list.

The assistants themselves are installed separately. Termspli shows how to install one when it is missing.

Features

• Several sessions per assistant, each in its own tab with its own terminal; switch tabs freely - the others keep working.
• Mix assistants: Claude Code, Codex, Kiro and Antigravity can all run at the same time on the same project.
• One agent per pane: split the window and start a different agent - or a different project - in each pane.
• Expand any session to the whole window, then put it back in the sidebar.
• Remote tasks started from your phone appear as tabs here too.

How to use

1. Open a project in Code and click AI in the editor's top bar.
2. Choose an assistant tab, then New session - or pick a past conversation and resume it.
3. Start another session the same way; each one gets its own tab.
4. Click a tab to watch that agent; click × to end it.
5. For agents on different projects, use Split right / Split down and start one in each pane.

Tip: give parallel agents separate files or separate branches, so they don't edit the same lines at once. Review the result in Source control before committing.

Features

• Project tree with file-type icons, sizes and dates; large folders load only when opened.
• Tabs with their own undo history, so switching files never forces a save.
• Quick open (⌘⇧O), find in all files (⌘⇧F) and go to definition (F12).
• Source control: changed files, side-by-side diffs, commit, branch switching and push.
• Format code with the project's own formatter (⌥⇧F), and compare any two files.
• Previews for Markdown, images, PDF, Word and spreadsheets; a built-in image editor.
• Local history keeps earlier versions of every file you save.
• Saved project commands to run, build or test the project in one click.

How to use

1. Press ⌘⇧E and choose a recent project - or open a new folder.
2. Click a file in the tree to open it in a tab; edit and press ⌘S.
3. Open Source control to see what changed; click a file to view its diff.
4. Write a message, commit, then click Git push.
5. Press F1 (or Keys) at any time to see every editor shortcut.

Features

• Secret scanning - keys and tokens are found and shown redacted.
• Sensitive files tracked by git, such as private keys and database dumps.
• Configuration rules for Dockerfiles, Compose files and GitHub Actions.
• Dependency check against the open OSV vulnerability database (only package names and versions are sent, and it can be turned off).
• AI review with Claude Code, Codex, Kiro or Antigravity for an overview and further findings.
• Fix selected: the AI changes only what you tick; every touched file is saved to local history first.

How to use

1. Open a project in the Code editor and click Code Review.
2. Choose the AI engine and which scanners to run, then click Run review.
3. Select a finding to read its evidence and the suggested fix.
4. Tick the findings you want fixed and click Fix selected.
5. Check the changes in Source control before you commit.

Features

• Multi-line commands with a title and an optional working folder.
• Run opens a new pane, so a long-running server never takes over your current shell.
• Insert types the command into the active pane for you to edit first.
• Sorted by use - the command you run most sits at the top.
• Stored as plain JSON you can read or back up.

How to use

1. Press ⌘⇧R and click New command.
2. Give it a title, paste the command, and optionally set the folder to run in.
3. Save with ⌘↩.
4. Select it and click Run - or double-click the title.

Features

• One list per project, created automatically, and named lists for everything else.
• Saved instantly - nothing is ever “unsaved”.
• Open a project's tasks straight from the code editor.

How to use

1. Click Tasks and choose a project or list on the left.
2. Type in Add a task and press Enter.
3. Tick a task when it's done.

Features

• Workspaces, environments, collections and folders to keep requests organised.
• Variables at workspace, environment or collection level, used as {{name}}; the resolved URL is shown under the address bar.
• Params, headers, auth, body and scripts for each request.
• Readable responses - a JSON tree, headers, cookies, status and timing.
• Import existing collections, and run a whole collection in one go.

How to use

1. Press ⌘⇧A, choose a workspace and an environment at the top.
2. Click New to add a request, or Import an existing collection.
3. Pick the method, enter the URL, and fill in headers and body.
4. Click Send and read the response below.

Variables are stored as plain JSON on your machine. Mark a variable secret to mask it on screen.

Features

• Virtual users with ramp-up and duration, like JMeter without the jargon.
• Chained steps: take a value - such as a login token - from one response and use it in the next request.
• Targets: web pages, HTTP APIs and databases.
• Live charts and a final report with percentiles, throughput and errors.

How to use

1. Click Testing and create a plan.
2. Add steps; on a step, choose the value to extract for the next one.
3. Set users, ramp and duration.
4. Click Start run and watch the results.

Test responsibly: run load tests only against systems you own, ideally a staging copy.

Features

• Photo: menu bar, options bar, toolbox, rulers and a dock of History, Swatches, Properties and Layers.
• Whiteboard: ⌘Z / ⇧⌘Z to step through strokes, ⌘+ / ⌘− to zoom.
• Diagrams: connections are attached to shapes, not to points, so the picture stays right as you rearrange it.

How to use

1. Click Design and choose Photo, Whiteboard or Diagram.
2. Pick a tool from the left column; set its options in the bar above.
3. In Photo, right-click or hold a tool to see the related tools in its group.
4. Save or export from the File menu (or Save… on the whiteboard).

Features

• JSON: validate, format, minify, sort keys and browse as a tree.
• JSON compare: see exactly what differs between two documents.
• Encode: Base64 (standard and URL-safe), percent-encoding, hex, query strings, and JWT decoding with readable expiry times.
• Hash: SHA-256, HMAC-SHA-256 and PBKDF2 with salt and iterations.
• Beautify: re-indent minified HTML, CSS, JavaScript and TypeScript.
• Identifiers: generate UUIDs and other IDs; Timestamp: convert Unix time to dates and back.

How to use

1. Click Utils and choose a tool on the left.
2. Paste your input on the left side.
3. Click the action you need; copy the result with Copy.

Private by default: the panel remembers which tool you used last - never what you typed.

Features

• SFTP and FTP servers saved with host, user and starting folder.
• Two listings - local on the left, server on the right - so you can see what differs.
• Transfer queue for uploads and downloads with progress.
• Edit in place: open a server file, save, and it is uploaded back.
• Passwords are kept in the operating system's keychain, never in a settings file.

How to use

1. Click Remote, then + Server and fill in the connection details.
2. Select the server to connect.
3. Choose a local folder on the left (you will be asked once to allow it).
4. Drag files between the sides, or double-click a server file to edit it.

Features

• Full web pages - sites that refuse to be embedded elsewhere still load here.
• Several pages side by side or stacked, each with its own address bar.
• Screenshots of a page, taken by the browser itself.

How to use

1. Click Browser and type an address.
2. Click Split to add a second page beside or below.
3. Use Screenshot to save what a page shows.

Features

• Start and stop the Docker engine.
• Every container listed - run, stop, clone or delete.
• Reclaim disk space from unused images and caches without losing a container.
• Edit Docker Desktop's settings, such as memory and CPUs, from here.
• Clear warnings before anything destructive happens.

How to use

1. Click Docker; start the engine if it isn't running.
2. Use ▶ / ■ on a row to run or stop a container.
3. Click Clean up to reclaim disk space - no container, running or stopped, is removed.

Features

• Four platforms: GitHub Actions, GitLab CI, Bitbucket Pipelines and Jenkins.
• Guided setup in the order questions come up: source, build, checks, deploy, after-deploy checks.
• AI help to work out how your project builds.
• Always a pull request - nothing is pushed to your main branch.
• Secrets stored in the platform's own secret store; a one-time deploy key for servers.
• Runs view showing what the platform did with each pipeline.

How to use

1. Click CI/CD, then + New, and choose the project.
2. Sign in to the platform (GitHub uses your gh login; others take a token).
3. Work through each section and review the files and secrets at the end.
4. Click Publish - a pull request opens on the platform for you to merge.

Features

• Transport: HTTPS, redirects and HSTS.
• Security headers, cookie flags and CORS policy.
• Version disclosure and exposed files such as .env or .git.
• Each finding has a severity, the evidence, and how to fix it - and “Good” notes show what is already right.
• AI explanation with Claude Code, Codex, Kiro, Antigravity or an API key.

How to use

1. Click Security, then + Add and enter the site's address.
2. Tick I own this site or have permission - the scan will not run without it.
3. Click Scan and read the findings.
4. Click Explain with AI for a ranked list of fixes.

Read-only and gentle: about ten ordinary requests per scan. It never tries to break in or overload the site.

Features

• Entry types: SSH, Login, Plain text and File (such as a private key, up to 10 MB).
• Two-step unlock: your passcode plus a code emailed through your own mail server.
• Folders, drag-and-drop, and sorting by name, last used or last changed.
• One-click SSH: connect in a new terminal pane, using a stored key if you like.
• Google Drive sync to a termspli folder; the newest copy wins and backups are kept.

How to use

1. Click Vault and create a vault - or restore one from Google Drive.
2. Choose a passcode and set up the email for unlock codes.
3. Unlock with the passcode and the emailed code.
4. Click + Entry, pick a type and fill it in.
5. On an SSH entry, click Connect in new pane.

Keep your passcode safe. The vault is encrypted with it, so it cannot be recovered without it.

Features

• Works with Claude Code, Codex, Kiro and Antigravity.
• No open ports. Your computer dials out to a small relay on your own server; the relay only passes sealed messages it cannot read.
• Three relay options: PHP for shared hosting, Node.js, or a free Cloudflare Worker.
• Passcode on every start and stop; five wrong tries lock the connection.
• Android app paired by QR code, with a notification when a task finishes.
• Status first - the output is shown only when you ask for it.

How to use

1. Open Settings → Operate remotely and choose a relay type.
2. Click Server files… and upload them to your server (one time).
3. Enter the relay address, add it to the allowed list, and choose a passcode.
4. Scan the device-link QR code with Termspli Remote, or open the link in a browser.
5. Pick a project and an assistant, type the prompt and passcode, and tap Start.

Features

• 13 themes, light and dark, applied to the app, the editor and the terminals.
• Appearance: fonts, sizes, cursor, colours - globally or per pane.
• Toolbar with icons or names; show or hide pane headers.
• New Window for a second, independent workspace.
• Translation with your own Google Cloud, DeepL or LibreTranslate key.
• Export and import all settings in one encrypted file.
• System info about your computer at a glance.

How to use

1. Click ⚙ in the toolbar and choose Theme or Appearance.
2. Choose Toolbar: show names or show icons to suit your screen.
3. To move to a new computer, choose Export settings, set a password, and Import settings on the other machine.