Routescan documentation
Routescan is a blockchain data platform for applications, analysts and AI agents. Read indexed transactions and balances, inspect historical state, and explore on-chain activity across EVM networks.
Routescan indexes active networks including Ethereum, Avalanche C-Chain, Mantle, and Blast. Service coverage varies by chain: check the supported-chain catalog before choosing an endpoint.
Install the CLI to sign in from your terminal, discover chains and make your first request. Or start with HTTP and use the API directly.
Choose a product
Install the CLI
Install rts on macOS or Linux with one command. The installer verifies SHA-256 checksums over HTTPS on both platforms. macOS releases are not yet Developer ID signed or notarized by Apple. For Apple-signed releases, the installer also checks the developer identity and notarization using system tools. No Python, GPG or Cosign is required for installation. The installer also adds the rts skill for Claude Code and Codex; start a new agent session to load it. Set RTS_NO_AGENT_SKILLS=1 on the installer process to skip skill installation.
Apple Silicon and Intel Macs · Linux arm64 and x64
For an independent signature check before installation, see Verify CLI downloads.
Check the installation
Open your terminal and check the installed version. A successful installation prints the rts version and platform.
Use with an agent
The installer adds a skill that teaches Claude Code and Codex when to use rts, how to discover Archive RPC endpoints and how to authenticate with an API key or device key. Start a new agent session, then ask it to use rts. If you copied the binary manually, install the skill with this command. Use --client claude or --client codex to choose one client; the default installs both. Customized existing skills are preserved.
Sign in
Sign in through the console in your browser. In a local desktop terminal, rts opens the browser automatically. Over SSH or on a headless server, open the printed link on your computer. Compare the authorization code and requested permissions before approving access. On the authorization page, choose Data access, Full access or Custom before approving. Full access is selected by default and includes account, API key and billing management. Choose Data access or Custom to restrict access. You do not need to copy an API key.
Preselect permissions
Login preselects every supported permission by default. Use a comma-separated list to request less access for a specific task; --permissions full explicitly selects the default. Review and adjust the selection on the authorization page before approving. A successful login replaces the credential saved on this device; it does not revoke previously issued keys.
Check access
Verify your saved credential and its permissions before making a request.
Find a chain
List the available networks and Routescan chain IDs. This catalog is public and can also be viewed before signing in.
Discover Archive RPC endpoints
List endpoint templates and authentication instructions for direct JSON-RPC calls. The CLI checks RTS_API_KEY or ROUTESCAN_API_KEY from the environment first, then the nearest project .env.local or .env, then the saved device key. It verifies rpc:read and reports a fallback to the device when needed. The output contains only a credential reference, never the secret. Agents should read that reference into memory and authenticate requests without printing the key or putting it in command arguments. The MCP RPC tool has separate method restrictions.
Read blockchain data
Request five indexed mainnet transactions across the supported chains. Replace all with a Routescan chain ID from the catalog to query a specific chain.
Manage Streams pipelines
List the chains and datasets a pipeline can read, check a definition, then create the pipeline and follow its messages. validate only reads and exits with code 1 when the definition is not valid; commands that change something need --yes. Commands that act on one item take its name as the first argument. The Streams documentation describes the definition format and every streams command.
Explore commands
Use the JSON command catalog and per-command help to discover the interface; per-command help lists flags, query parameters and examples. Output is JSON automatically when piped or captured; --json makes that choice explicit. Use --json=false for text. --query can be repeated. Errors and progress stay on stderr; in JSON mode they are JSON lines. Failures include an exit code and recovery guidance when available: 2 for invalid input, including input errors rejected by the API, 3 for failed authentication or a missing permission, 4 for rate limits and 5 for network errors. Use --no-input to disable authentication prompts and automatic browser launch.
Update
Shell installations check for updates in the background once an hour. Set RTS_NO_AUTO_UPDATE=1 to disable automatic updates, or set RTS_VERSION=X.Y.Z when installing to pin a version. Run rts update to update manually. A binary copied manually does not include this updater; run the installer to use it.
Sign out
When you want to disconnect, sign out to revoke the saved credential and remove it from this device. Credentials supplied through RTS_API_KEY are not revoked by this command.
Troubleshooting
Command not found
Open a new terminal and retry rts --version. If it still fails, check that ~/.local/bin is on PATH, or the directory you selected with RTS_INSTALL_DIR.
Browser did not open
Over SSH or on a headless server, this is expected: open the printed URL in a browser on your computer. Compare the code and permissions, approve access, then return to the terminal on the server. The CLI waits for approval automatically. To prevent automatic browser opening in any environment, use --no-open.
Coming next
Querybase — coming soon. A serverless SQL database of decoded chain history.
Manage your workspace
Use the console to manage access and review usage. Browse supported chains for API coverage and Archive RPC for node endpoints.
Questions? Contact us