Backport release/v6.7: Upgrade UCI AI review to latest and allow review of PRs from seidroid

[seidroid]

Backport of #4059 and #4067 to release/v6.7.

[cursor]

PR Summary

Low Risk
Only GitHub Actions workflow pins and bot allowlisting; no application or runtime code changes.

Overview
Backports CI updates so AI Assistant and AI Review reusable workflows pin sei-protocol/uci from v0.0.19 to v0.0.21 (new commit on both uses: and uci-ref).

AI Review also passes allowed-bots: '["seidroid[bot]"]' so automated backport PRs opened by seidroid can run through the same review flow as human-authored PRs.

^{Reviewed by Cursor Bugbot for commit a0cbdbc. Bugbot is set up for automated code reviews on this repo. Configure here.}

[seidroid]

The automated review did not complete; see the failing AI Review check for details.

[github-actions]

The latest Buf updates on your PR. Results from workflow Buf / buf (pull_request).

Build	Format	Lint	Breaking	Updated (UTC)
✅ passed	✅ passed	✅ passed	✅ passed	Sep 1, 2026, 12:20 PM

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 60.30%. Comparing base (5968480) to head (a0cbdbc).

Additional details and impacted files

@@               Coverage Diff                @@
##           release/v6.7    #4063      +/-   ##
================================================
- Coverage         61.34%   60.30%   -1.04%     
================================================
  Files              2163     2064      -99     
  Lines            188889   177206   -11683     
================================================
- Hits             115868   106869    -8999     
+ Misses            62286    60545    -1741     
+ Partials          10735     9792     -943     

Flag	Coverage Δ
sei-db	69.80% <ø> (ø)
sei-db-state-db	?

Flags with carried forward coverage won't be shown. Click here to find out more.
see 99 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[masih]

@seidroid review

[seidroid]

Straightforward backport that bumps the UCI reusable-workflow pins and adds allowed-bots: '["seidroid[bot]"]' so AI review runs on automated backport PRs; both workflow refs remain SHA-pinned and the trigger set is unchanged. No blocking issues — only a note that this backport pins v0.0.21 while the source PR pinned v0.0.20, and that allowed-bots was added to the review workflow but not the assistant workflow.

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

Blockers

• None at the file/PR level.

Non-blocking

• [suggestion] This backport deviates from its source PR (#4059), which pinned UCI v0.0.20 (68010d0ab76554da91c320511fb14a7af1af4245); here both workflows are pinned to v0.0.21 (93ebad0186c8b71a4036fd0915ffe2ec44283d4b). That leaves release/v6.7 running a UCI revision main does not use, so the two branches' AI review/assist behavior can drift and a regression in v0.0.21 would surface first on the release branch. Consider bumping main to v0.0.21 as well (or backporting v0.0.20 verbatim) so the pins stay in lockstep.
• [suggestion] allowed-bots is added only to ai-review.yml, not to ai-assist.yml, which keeps allowed-team as its sole gate. If that asymmetry is intentional (assist is comment-driven and seidroid[bot] does not issue assistant commands), it is fine as-is; if backported PRs are also meant to be usable with the assistant, the same input is needed there.