[taras]

Why

CI needs measured costs for the REPL include regression delivered in #972 so weighted test shards can account for it alongside the rest of the corpus.

What changes

Before: repl-include-pty.test.ts has no recorded weight and receives the conservative maximum-weight fallback for each runtime.

After: CI allocation uses its measured Deno, Node and Bun costs, together with refreshed per-file weights for the same corpus. This updates test-weights.json with the successful runner artifact byte for byte.

Generated or mechanical changes

The artifact comes from Measure test weights run 39556424450, attempt 1, on ubuntu-latest. Its provenance remains source commit a4907b468777e4112f87ff80d26f0e692fa73e35, measured with Deno 2.9.5, Node v22.23.3 and Bun 1.4.0. That source tree equals the merged #972 tree and this follow-up's parent tree.

How to verify it

• The committed 74,638-byte file matches the downloaded artifact exactly. Validation finds complete coverage for Deno 489, Node 381 and Bun 380 files, including the REPL include regression in each.
• deno task test scripts/tests/test-weights.test.ts scripts/tests/measure-test-weights.test.ts passed all 22 steps. Both original reviewers independently returned PASS for signed head b403958b23e4bc9e486525fd460aa6e280bc624d.
• Required CI run 39582551750 completed successfully, including green, at that same head.

Scope

Only test-weights.json changes. Test selection, exclusions, runner commands, shard counts, dependencies and product behavior keep their existing contracts. This PR applies measurement data; it adds no measurement mechanism or feature implementation.

Scope confirmation

• Every changed file supports the purpose described above.
• Unrelated cleanup and formatting changes are excluded.
• Generated or mechanical changes are clearly identified.
• The description matches the final diff and test results.

[github-actions]

PR #973: chore: measure CI cost of the REPL include regression

1 files, +1252 / -1110

Scope

🔴 PR has 2362 lines changed. Split into focused PRs.

🟡 2362 lines changed. PRs under 400 receive more thorough review.

Structural

✅ No structural bloat detected.

Slop

✅ Slop indicators look low.

Static Analysis

✅ Oxlint found no issues.

Correctness

No extraneous code patterns detected.