Quickstart: Tempr CLI
Last updated
Install the CLI, sign in with your Tempr license key, add a provider key, and run your first agent turn from the terminal — usually under two minutes.
-
Install the CLI
The CLI is a self-contained build for Windows, macOS, and Linux (x64 and Arm64), so there's no runtime to install first. On macOS and Linux:
curl -fsSL https://temprhq.io/cli/install.sh | shOn Windows, in PowerShell:
irm https://temprhq.io/cli/install.ps1 | iexOr download a build directly and put
tempron yourPATH; each download has a.sha256beside it. -
Sign in
Sign in once. Choose your browser: the CLI shows a code and opens the Tempr portal, where you approve it — from any device, so it works over SSH too. Or paste the same license key you use in the IDE:
tempr authRunning
temprfor the first time does this for you, then offers the next two steps. -
Add a provider key
Tempr calls every model with your own key for its provider, such as OpenAI, Anthropic or Google. Add one, pick the provider from the list, and paste the key at the hidden prompt. Tempr checks it with the provider before saving it:
tempr keys addKeys you've already added in the IDE or the Portal are there too;
tempr keyslists them. Azure, AWS Bedrock and custom endpoints need more than a key, so they're set up on the Portal's Provider keys page. On a Team plan, your organization's admin manages keys. -
Pick a default model
Any model from your plan's catalog on a provider you've configured works.
tempr authoffers to pick one right after you sign in; otherwise list them, set a default, and override it per-prompt with-mwhen you want:tempr models tempr config set-model <id> -
Run your first turn
One-shot a task from your repo's root, or start an interactive session:
tempr "explain this repo" tempr
Three ways to run it
| Invocation | Behavior |
|---|---|
tempr "<prompt>" | One-shot. A full agent turn with real tool-calling — file read/write/edit, run_command, and semantic search — then exits. |
tempr | Interactive REPL. Persistent cross-session history, line editing, and tab-completion for /commands and @personas. |
<something> | tempr | Piped / headless. Reads the prompt from stdin, writes to stdout — composable like any other command-line tool. |
Personas and plain chat
The built-in specialist personas work on the command line with the same names as the IDE:
tempr "@debugger why does this fail"If you want a quick answer with no file access or commands, use ask mode:
tempr --ask "what's the difference between IActionResult and ActionResult?"Managing your setup
| Command | Does |
|---|---|
tempr config | Manage settings and your default model. |
tempr models | List the models your license can use. |
tempr keys | List, add, and remove your provider keys. |
tempr history | List, show, rename, fork, and delete saved sessions. |
tempr mcp | Add, list, and manage MCP servers. |
tempr doctor | Check connectivity and diagnose your setup. |
tempr update | Check for a newer release and print the command that installs it. |
tempr --help | Full command reference. |
The CLI runs on your existing Tempr Chat license — there's no separate subscription. The agent loop runs server-side, so a dropped connection doesn't lose an in-flight turn.
Next steps
- CLI command reference — every command, flag, REPL shortcut, and the agent tool set.
- Scripts, pipes & CI — headless patterns and the JSON event stream.
- Providers & BYOK — the supported-provider list behind your model choices.
- Agent modes & personas — the persona lineup and how delegation works.
- MCP servers — extend the agent with your own tools and data sources.