SLV Benchmark Skill

Benchmark and connectivity workflows for SLV.

Supported benchmark types

• shredstream
• grpc
• rpc

Input collection order

For benchmark requests, collect inputs in this order:

1. benchmark type
2. region
3. endpoint URLs

Region should be collected before execution because --region gives more accurate measurements for feed comparison.

geyserbench config generation

Use:

• ~/.slv/api.yml for the ERPC API key when available
• https://edge.erpc.global as erpc_url
• two endpoint URLs for side-by-side comparison

Example config:

[config]
region = "frankfurt"
erpc_url = "https://edge.erpc.global"
erpc_api_key = "api-key"
transactions = 10000
account = "pAMMBay6oceH9fJKBRHGP5D4bD4sWpmSwMn52FMfXEA"
commitment = "processed"

[[endpoint]]
name = "http://endpoint-1"
url = "http://endpoint-1"
kind = "shredstream"

[[endpoint]]
name = "http://endpoint-2"
url = "http://endpoint-2"
kind = "shredstream"

Planned CLI shape

A natural future command shape is:

slv check geyserbench <options>

This should mirror the current slv check grpc / slv check shreds style by passing arguments through to the installed benchmark binary.

API key handling

If ~/.slv/api.yml does not contain the required ERPC API key, instruct the user to get a free API key and configure it first.