docs(indexers): update Envio page to current HyperIndex setup

[Jordy-Baby]

Hey @GigaHierz, this refreshes the existing Envio page in the Celo docs so it matches how HyperIndex works today.

• Restructures the page to the order in AGENTS.md
• Updates the envio init walkthrough to the current CLI and adds the non-interactive commands
• Lists Celo (42220) with its HyperSync endpoint, and Celo Sepolia (11142220) through an RPC data source
• Removes links to envio.dev/explorer, which now redirects to the homepage, plus goerli and Alfajores references

Every command was run against the current CLI, every link returns 200, and mint broken-links reports none.

Lmk if you require anything further on our end.

🤖 Generated with Claude Code

[GigaHierz]

This is a real improvement and I want it in. I verified the substance rather than skimming it, and the claims hold up — including the ones that correct the existing page.

What I checked:

• Celo mainnet HyperSync is real at source, not just in your docs: packages/cli/src/config_parsing/chain_helpers.rs has Celo = 42220 inside the HypersyncChain subenum, so HyperIndex genuinely defaults to HyperSync there. GET https://celo.hypersync.xyz/height returns 200.
• Celo Sepolia really is absent from the network enum (grep 11142220 → 0 hits), so the RPC-data-source treatment is correct rather than a shortcut.
• The interactive walkthrough matches interactive_init/mod.rs and shared_prompts.rs verbatim and in order, at both 3.12.1 and 3.13.0. The non-interactive flags all parse, including the two easy-to-get-wrong ones — -c and -d work in those positions only because they're declared global = true.
• envio.dev/explorer does 308 to the homepage, so removing those links is right. All 20 links in the added content return 200.
• The old page's "JavaScript" handler language was wrong; "TypeScript (the default) or ReScript" matches the -l values.

Notably this removes a false claim we are currently publishing — "native support for Celo networks (testnet and mainnet)" — which is the main reason I'd like it to land.

Two corrections and one addition, then I'll merge.

1. The Go client is not on the same footing as the others

HyperSync is also available as a standalone API through the Python, Rust, Node.js, and Go clients

Your own clients page marks the Go client community maintained and work-in-progress. Listing it inline with the three supported ones sets the wrong expectation. Suggest:

...through the Node.js, Python, and Rust clients (a community-maintained Go client also exists)

2. The Celo Sepolia sentence is accurate but under-specified, in a way that will cost readers time

For networks without HyperSync, such as Celo Sepolia (chain ID 11142220), HyperIndex uses an RPC data source

True, but it reads as though only the data source differs. Celo Sepolia is absent from the network list entirely, so contract-import explorer won't work for it either — there's no explorer entry, only celo-alfajores and celo-baklava. A reader will try the first code block with a Sepolia address and get a confusing failure. One clause fixes it, and it also explains why the two code blocks use different subcommands, which currently reads as unmotivated:

...HyperIndex uses an RPC data source. Celo Sepolia is not in Envio's network list, so you must supply the ABI and RPC URL yourself rather than importing from a block explorer.

3. Please fold in the indexers overview

tooling/indexers/overview.mdx:18 still carries the old marketing blurb and doesn't link to this page. Once this merges the two pages disagree with each other, which is exactly the drift our style guide's one-fact-one-page rule exists to prevent. Same PR, please.

---

A note for your side, not a blocker: Envio's supported-networks page still lists celo-alfajores and celo-baklava. Alfajores is retired and Baklava is long gone — Celo Sepolia (11142220) is the current testnet. Worth adding it and dropping those two.

I've merged #2333 separately since it only links this page and stands on its own. Also flagging for the record that this is vendor-authored — that's fine and the claims check out, but it's worth having stated.

Re-request review when the three are in and I'll turn it around quickly.

[GigaHierz]

Friendly nudge — no rush, just making sure this hasn't fallen off your radar.

The three changes from my review are all small, and #2333 (the indexer-selection link) merged, so this page is now reachable from the scaling page while still carrying the stale "native support for Celo networks (testnet and mainnet)" line. Landing this replaces that with the accurate mainnet-HyperSync / Sepolia-RPC split, which is the main thing I want out of it.

If you'd rather I pushed the three edits to your branch myself, say the word and I will — happy either way.

[Jordy-Baby]

Hey @GigaHierz, yes please go ahead and push the three edits to the branch. Thanks for the thorough review. lmk if you require anything further on our end.

[GigaHierz]

Pushed the three edits to your branch as you offered — thanks. Also merged current main in, since the branch was based on 784daebd and had drifted behind the redirect work that landed since.

What I changed

1. Go client — now "Node.js, Python, and Rust clients (a community-maintained Go client also exists)", since Envio's own clients page marks it community-maintained and work in progress.
2. Celo Sepolia — added the clause: "Celo Sepolia is not in Envio's network list, so you must supply the ABI and RPC URL yourself rather than importing from a block explorer." That is also what makes the two different subcommands in the code blocks read as deliberate rather than arbitrary.
3. tooling/indexers/overview.mdx — replaced the old blurb, which still described Envio as "a modern, multi-chain EVM blockchain indexing framework speed-optimized for querying real-time and historical data", and added a link through to this page. Without that the two pages would have contradicted each other the moment this merged.

One thing worth flagging: the branch was behind far enough that broken-links --check-redirects failed on it with nine pre-existing redirect errors — all from before #2328 landed, none from your diff. Merging main cleared them. Nothing for you to do; noting it so the earlier red run does not look like it was yours.

$ npx mintlify@4.2.920 validate                        -> build validation passed
$ npx mintlify@4.2.920 broken-links --check-redirects  -> no broken links found
$ bash scripts/check-orphans.sh                        -> No orphan pages found

Approving. This replaces a claim we were publishing that is not true — "native support for Celo networks (testnet and mainnet)" — with the accurate mainnet-HyperSync / Sepolia-RPC split, which was the main reason I wanted it in. Merging once the check goes green.

Still worth doing on your side, unrelated to this PR: Envio's supported-networks page lists celo-alfajores and celo-baklava, both retired. Celo Sepolia (11142220) is the current testnet.