[kriszyp]

⊙ Problem

main fails Lint Source Code: resources/DESIGN.md is 1001 lines; the budget is 1000. Two merges each passed the design-doc check alone: fix: propagate residency read failures and refresh stub indexes (#3322) took the file to 997 lines, then Require a super user for X-Replicate-To … (#3084) added a 4-line note on top.

🤔 Is the requirement right?

Yes. The budget exists to force pruning as notes are added; the fix is a prune, not a higher budget.

💡 Solution

Tighten two wordy notes in resources/DESIGN.md, keeping every fact, symbol, test path and issue link: 1001 → 991 lines.

• The nested-selection bullet under Conventions (transformForSelect/transformEntryForSelect): the densest of the two; check the three issue links still read right.
• The RocksDB teardown note (ownedRocksStores, harper#3120)

🤔 Only 9 lines of headroom

Shrinking notes to one-line "enforced by <test>" statements would buy far more room. Decide per note in the quarterly prune-design-docs pass, not a fix for failing main. Expect this check to trip again within a few merges.

✅ Verification

• npm run check:design-docs passes on this branch and on its merge with current main (7ce7da60d).
• prettier --check resources/DESIGN.md passes.

Signed: Claude Opus 5.5

🤖 Generated with Claude Code

Release target: main — the diff changes only tests, docs or lockfiles
Related PRs: #3221 overlaps, #3126 overlaps, #3375 overlaps, 5 others independent

Origin — the dispatch brief this PR was written from

harper main is red: Lint Source Code at 0a030f3

main on HarperFast/harper went from green to red. Failing workflows: Lint Source Code at 0a030f3.

Find the merge that broke it and get main green. Walk that workflow's recent push runs on main oldest-to-newest to find the FIRST red head SHA and the PR it came from, then pull the failing test names from that run and from the current head. A failure on every matrix leg (Node version, Bun, uWS, Windows) is a regression; one leg is usually a flake.

Then either fix it or, when the culprit is a single merge and the fix is not obvious, open a revert and say so — main being green is worth more than the change being preserved. Record the fingerprint on the initiatives/ci-health board doc either way (prose row beginning with the test path, never a bare ref), and add your PR to initiatives/testing-and-deploy > "CI & build infrastructure" as a bare-ref line so it renders as a card.

_{Dispatch: task main-red-kriszyp_harper_0a030f35a_5684deb1 · queued by automation · ran by claude/opus/high · worker kzyp-xps-1}

_{Review-Coverage: authored=claude; ran=codex,gemini; adjudicated=domain; declined=cursor-grok,cursor-composer,cursor-kimi,cursor-muse,muse; rounds=2; full=1 @ b591a12}

_{Review-Attention: skim ~2m (decisions: condense-vs-shrink-to-enforcement) @ b591a12}

[gemini-code-assist]

Code Review

This pull request updates the documentation in resources/DESIGN.md to clarify the conventions around Resource.transformForSelect and Table.transformEntryForSelect, as well as the teardown process for RocksDB database wrappers. The changes improve readability and update references to open issues. There are no review comments to assess, and I have no feedback to provide on these documentation updates.

[kriszyp]

Heads-up for the budget count: #3619 Fix collection, filter and GraphQL reads of a @relationship into another database adds one note to resources/DESIGN.md (4 lines: heading plus one paragraph, "A related-table read uses the related table's own database link"). Whichever lands second will need the other's lines in its count.

_{🤖 Generated by Claude Code (Claude Opus 5.5); posted via @kriszyp.}