Fix Base/Extended checkmarks in supported networks table to reflect EVM block model

[brandonleekramer]

Summary

The Substreams and Firehose columns on the Supported Networks table were labeling their checkmarks as Base (single check) vs Extended (EVM only) (double check) — a claim about the network's EVM block model — but the underlying logic derived that distinction from the number of providers serving the network, which is unrelated. As a result the marks disagreed with the registry's firehose.evmExtendedModel field.

Examples of the mismatch (registry main):

• Avalanche — evmExtendedModel: false, but 2 Substreams + 2 Firehose providers → showed a double check ("Extended"), should be single ("Base")
• HyperEVM — evmExtendedModel: true, but only 1 provider each → showed a single check ("Base"), should be double ("Extended")

Change

Replace the two provider-count functions (getSubstreamsSupportLevel / getFirehoseSupportLevel) with a single getFirehoseSubstreamsSupportLevel that reflects the block model. Substreams and Firehose share one signal (both are powered by the same Firehose block data, and the table's legend already groups them):

• none — no Firehose or Substreams provider is serving the network
• basic (✓) — base blocks + at least one Firehose or Substreams provider
• full (✓✓) — extended (EVM) blocks (firehose.evmExtendedModel) + at least one Firehose or Substreams provider

Both columns render this same combined mark.

Behavior notes

• Networks with the extended block model but no providers serving the data yet (e.g. Katana) correctly remain off the table, since they have no indexing service.
• ~22 networks shift their Substreams/Firehose marks under the corrected rule; row inclusion is unchanged.

Testing

• pnpm typecheck passes (no new type errors).
• Verified against the live registry aggregate (v0.7.121): Avalanche → ✓, HyperEVM → ✓✓, Katana → not shown.

🤖 Generated with Claude Code

https://claude.ai/code/session_01NEfkbiG5zT1qGbYytN2uUW