Announcing the Tailstream Client: Fast Terminal Access to Live Logs

Today we are releasing the open source Tailstream Client, a dedicated command-line interface for engineers who want fast, scriptable access to their Tailstream streams. The CLI mirrors the Tailstream web experience with OAuth authentication, interactive log navigation, and the same precision filtering used across our platform—packaged in a binary that fits neatly into terminals, automation, and incident response workflows.

Why we built a CLI

Tailstream streams frequently live inside tmux panes, SSH sessions, and CI jobs. We wanted the Tailstream Client to deliver the same real-time visibility you expect from the dashboard without ever leaving the shell. Engineers can install pre-built binaries for Linux (amd64/arm64) and macOS (Intel/Apple Silicon), or compile from source with the provided build.sh script. Every build ships with production-ready defaults so you can authenticate and stream logs against your Tailstream organization immediately.

Secure, low-friction authentication

The client uses the Tailstream OAuth 2.0 device flow so you never have to paste tokens or share credentials. Run tailstream-client --login, approve the request in your browser, and encrypted credentials are stored in ~/.tailstream-client.yaml. Tokens rotate automatically, --logout clears local state, and automation can still supply --token explicitly when a non-interactive workflow requires it.

Interactive log navigation built for speed

Launching tailstream-client --from "-1h" opens the default interactive viewer. Familiar Vim-style controls (j/k to move, g/G to jump, d/u to page) make scanning streams second nature. Expand entries inline with Space to inspect structured payloads, or fall back to --no-interactive and --json when piping output into scripts.

Cursor-based pagination keeps results fast even across large time ranges. Combined with stream-aware defaults, quick stream switching, and syntax highlighting for log levels, the CLI provides context that is easy to keep open beside editors and monitors.

Precision filtering and search

Tailstream Client respects the same filtering semantics as the hosted UI. Range queries accept relative or absolute timestamps (for example --from "-30m", --from "2025-10-01" --to "2025-10-02"). Server-side filters such as --level ERROR and --method POST combine cleanly with repeated --search flags for client-side fuzzy matching. Each search updates in real time as new records arrive, allowing you to triage incidents without re-running commands.

Configuration that adapts to your workflow

Preferences—including your last selected stream and time range—live alongside credentials in the YAML config file so recurring queries reuse sensible defaults. Because the project is MIT-licensed Go code, you can extend the CLI, embed it in internal tooling, or fork features to suit specialized pipelines.

Get started

1. Download the latest release for your platform from the GitHub releases page.
2. Mark the binary executable and move it onto your PATH.
3. Run tailstream-client --login to authorize via the device flow.
4. Launch tailstream-client --from "-1h" (or any preferred range) to stream logs interactively.

We are excited to see how you adopt the Tailstream Client alongside the Tailstream agent and UI. Issues and pull requests are open—share feedback, file feature ideas, and help us keep the terminal experience as capable as the dashboard.