[alexander-sei]

What is the purpose of the change?

Adds new content for PLT-837: an EVM parity matrix and plan for sei-js. Both were first written on the sdk-evm-parity branch of sei-js and reverted there. This re-lands them in sei-docs, which is now the canonical docs home, and re-scores every row against the current docs and packages.

The existing EVM compatibility page compares the Sei network with Ethereum. The new matrix answers a different question: which viem, wagmi, and ethers workflows these docs and @sei-js actually cover, and which gaps belong to sei-js rather than to the network.

Describe the changes to the documentation

• evm/sei-js/evm-parity.mdx (new): scores 24 developer workflows once each for viem, wagmi, and ethers, with links to the pages that cover each one. Gaps that sei-js owns are listed apart from Sei network differences. The current score is 54 of 71 scored cells (76%). Most open gaps are missing wagmi examples. No cell is marked Verified, because no automated check runs these workflows against a Sei node yet.
• evm/sei-js/evm-parity-plan.mdx (new): the goals (95% of scored cells, no evm-compat issue open longer than 14 days), principles, scoring rules, and how compatibility reports are classified.
• docs.json: adds both pages to the sei-js Library group.
• evm/sei-js/index.mdx and evm/evm-parity/evm-compatibility.mdx: a link to the matrix on each page, plus a link to the new issue template under Community and support.

Target audience: developers who arrive with viem, wagmi, or ethers and want to know what works on Sei, and sei-js maintainers who keep the matrix current.

Notes

• Merge after docs: restore the EVM compatibility issue template and SLA (PLT-837) sei-js#361. These pages link to the issue template and SLA doc that PR adds.
• The original plan's phased roadmap (conformance harness, SDK work) is left out, because AGENTS.md rules out speculating about unreleased work. The Phase 2 JSON-RPC conformance harness should be tracked as its own ticket.
• Checked locally with vale (no findings on the new pages), typos, mint broken-links --check-anchors on the changed pages, the docs.json page and redirect checks, and a mint dev render.

Made with Cursor

[mintlify]

Preview deployment for your docs. Learn more about Mintlify Previews.

Project	Status	Preview	Updated
sei-docs	🟢 Ready	View Preview	Nov 11, 2026, 1:24 PM

💡 Tip: Enable Automations to automatically generate PRs for you.

[cursor]

PR Summary

Low Risk
Documentation-only additions and nav/cross-links; no runtime or application code changes.

^{Reviewed by Cursor Bugbot for commit 08b24e7. Bugbot is set up for automated code reviews on this repo. Configure here.}

[seidroid]

This PR adds a well-structured EVM parity matrix and plan for sei-js. I recounted every cell, and the tallies are right: viem 21/24, wagmi 12/23, ethers 21/24, 54/71 = 76%. All internal links and anchors resolve, the network facts match evm-compatibility.mdx and reference.mdx, and the copy follows AGENTS.md terminology. I found no blockers; the notes below cover merge order and upkeep.

Findings: 0 blocking | 6 non-blocking | 2 posted inline

Blockers

• None at the file/PR level.

Non-blocking

• Merge order: the issue-template link (?template=evm_compatibility.md) and the EVM_COMPAT_SLA.md link depend on sei-protocol/sei-js#361. Merge that PR first, or these links will 404 or fall back to a blank issue form.
• Upkeep: the score table and the per-section matrices hold the same counts in two places. A short maintainer note (for example an HTML comment) saying to recount the summary table whenever a cell changes would help prevent drift.
• The second-opinion passes produced no output: codex-review.md and cursor-review.md were both empty, so this review is the only pass.
• REVIEW_GUIDELINES.md was empty, so I reviewed against AGENTS.md conventions only.
• 2 suggestion(s)/nit(s) flagged inline on specific lines.

[seidroid]

[nit] Slightly over-broad: the linked Pending state section says eth_getTransactionCount(address, "pending") does count the sender's mempool transactions. Only eth_getBlockByNumber("pending") returns the latest block. Consider saying "The pending block returns the latest committed block" so readers don't assume pending nonces are unsupported.

[seidroid]

[nit] The pinned sei-chain v6.6.1 will go stale as soon as the CI test is bumped. Consider linking to the test or workflow in sei-js instead of hard-coding the version, in line with the AGENTS.md guidance against hard-coding values that change.