Choose how to send data

You can send data to Squasher in several ways, and you can use more than one in the same project. Errors from an SDK and logs from your host then appear side by side in the project.

Compare the ways

Way	Use it when	What you get
SDK	Your app is JavaScript or TypeScript.	Errors with stack traces, user context, Web Vitals, and session replay.
Log drain	Your host can forward logs, and you want no code change.	Logs. Error-level logs become issues.
OpenTelemetry	You use another language, or you already use OpenTelemetry.	Traces, logs, and metrics. Error spans and error-level logs become issues.
HTTP API	You want to send JSON from your own code or pipeline.	Logs and errors in a format that you choose.

You can start with a log drain for fast coverage, then add an SDK where you want stack traces and user context.

Pick a guide for your stack

Stack	Guide
JavaScript and TypeScript	Next.js, Node.js, Browser, Cloudflare Workers, Pino, Winston, Browser logger
Python, Go, and Rust	Python, Go, Rust
Java, .NET, Ruby, PHP, and Elixir	Java, .NET, Ruby, PHP, Elixir. These use OpenTelemetry.
Hosting platforms	Vercel, Railway, Render, Fly.io, Heroku, Cloudflare
Cloud logs	AWS CloudWatch, Azure Monitor, Google Cloud
Servers, containers, and clusters	Host metrics, Docker, Kubernetes
Log pipelines	Vector, Loki, Splunk HEC, HTTP API
Any OpenTelemetry SDK or Collector	OpenTelemetry
AI agents and model calls	Agent SDK, OpenRouter
An app that already uses a Sentry SDK	Sentry SDKs

The npm packages under @squasher-ai/ are for JavaScript and TypeScript. There is no Squasher package on PyPI or crates.io. For other languages, follow the guide for your language.

Set up a log connector

A log connector gives a log source its own destination URL and key. Hosting platforms, log pipelines, and the HTTP API use connectors.

The dashboard has no page for log connectors yet. Create and manage them with the CLI, the API, or the MCP server.

Find the connector key

List the connector types. Each has a key, for example vercel_drain, railway_drain, otlp_http, http_api, or vector.

squasher log-connectors catalog

Create the connector

onboard creates the connector and returns the setup steps for that source, with the destination URL and the headers to use.

squasher log-connectors onboard --project "$SQUASHER_PROJECT_ID" --connector-key vercel_drain

Send a test log and verify

Configure the source with the values from the last step. Send one error-level log, then check the delivery:

squasher log-connectors verify --project "$SQUASHER_PROJECT_ID" <connector_id>

The other subcommands are list, get, update, delete, and rotate-key. For the API, see Log connectors. Before you send a large volume of logs, read Design queryable logs.

Get started

Guides

Choose how to send data

Compare the ways

Pick a guide for your stack

Set up a log connector

Next steps

Quickstart

Troubleshooting

​Compare the ways

​Pick a guide for your stack

​Set up a log connector

​Next steps

Quickstart

Troubleshooting

Compare the ways

Pick a guide for your stack

Set up a log connector

Next steps