Seer in a Nutshell

Seer facilitates development of Solana programs by enabling transaction debugging.

This document demonstrates how it works and shares implementation details.

Seer is a Drop-in Replacement

Before releasing, developers test their programs using solana-test-validator in their integration and e2e tests.

When an error occurs, debugging is limited to logs that must be declared in code beforehand, creating a loop: log → test → log.

Seer brings essential developer tooling to Solana: a debugger. When you run into an error, you don't need to rely on logs to find the cause - you can see the full call stack.

Your code doesn't need to be changed. Seer is a drop-in replacement for solana-test-validator.

How It Works in Practice

1. When you're ready to test your application, run the CLI command:

   seer run

2. Your application files (.rs and .so) are loaded to Seer Backend to add execution context to the Solana Validator instance.
3. You're ready to run your tests, e.g.:

   cargo test

4. You can see your transactions with their call stacks in the Seer Dashboard in your browser.
5. Want to test a new version of your code? Return to step 1.

Example of transaction trace:

Under the Hood

Seer builds a debug context using your application files (.rs Rust sources and .so compiled Solana programs) and starts a new Solana Validator instance with it. When the Solana Validator receives a new transaction, it uses the debug context to extract all significant steps and build the call stack.