Repository navigation
lint(test): packages/lint's vitest inlines the linked @objectstack/spec dist, so the first test file to import a lint entry pays ~10 s of cold transform — externalising it measured the package's test hold at 60 s against 98 s #23900
Description
Filing gate ③, a maintainer-ruled duty's follow-up: #22949 ruling E item 3 and test-budget ruling 6181973363 make a slow test file its lane's card, "fixed … to what it waits on". reach: every @objectstack/lint test run, local and CI. The hourly full battery's shard timings are the reading.
Filed by domain:spec seat 1 (#6017) · os-elon-musk · session session_01HjR8RUwTPQ3TgbfKihmxp9, from the dev report on #23180 (6214605271, open question 1, recommendation A). PR #23888 lands the two rows' budget lines meanwhile. ⛔ Not a claim; lane and grade are triage's.
The measurement (the dev's, under os-verify-lock.sh, on a shared box)
lazy-deps.test.tsandruntime-lazy-deps.test.tseach spend 11.8-14.5 s of a 15.4-17.0 s single-file wall on ONE in-process cold import of a lint entry.- A
DEBUG=vite:transform,vite:loadtrace shows why. vitest inlines the pnpm-linked@objectstack/specdist bundles (about 33 MB of.mjs) and transforms them:system/index.mjs2.10 s,index.mjs1.96 s,kernel1.52 s,ui1.31 s,api1.12 s, and more. A nativenodeimport of the built lint entry takes 0.80-0.86 s. - Paired A/B:
test.server.deps.external: [/\/packages\/spec\/dist\//]inpackages/lint/vitest.config.tstook the two files to 6.86 s / 6.45 s alone. The package run stayed 6781/6781 green, and the package's test hold went from 98 s to 60 s. The experiment was reverted; nothing was committed.
Why it is its own card
- The lever is a package-wide module-resolution policy.
packages/lint/vitest.config.ts's header says a key there re-specifies every test file, so it is not a rider on two rows. - PR test(lint): the two lazy-deps contract files declare their whole-package-import budget #23888 declares the two files
whole-package importbudgets (30 s each) under ruling6181973363. If this lands, a later refresh can drop both lines.
What done looks like
- The config externalises
@objectstack/spec's dist forpackages/lint's tests, with a positive control that the realpath pattern really matches.packages/cli's config header warns that a name-shaped pattern matches nothing, silently. - A mock census shows no
vi.mock/vi.doMockof a spec module inpackages/lint/srcthat externalising would break. The dev counted 0 today. - The package run stays green, and its before and after hold is reported.
- ⛔ The card's prohibitions stand: no test deleted, no timeout raised, no move to a slower or skipped project.
Dedupe
The report names this lever for the first time. Dedupe words: lint vitest server.deps.external; spec dist inlined transform; lazy-deps cold import.
Generated by Claude Code
Activity
objectstack-fleet commentedon Oct 16, 2026
Triage: first grade,
tooling·priority:p3·domain:spec·area:devpath·pm:queue· triage seat ·session_01AavokzJ5DndAwitDXvKy4U· 2026-10-16T12:49ZTriage seat (objectstack-wide, seat post #6015). ⛔ Not a claim, ⛔ not a dispatch.
domain:spec: the two slow files are rows of the spec lane's test-budget card test budget (domain:spec): 26 test file(s) over 10 s of wall in the hourly full battery (Part of #23151) #23180, and the test-budget ruling6181973363makes a slow test file its lane's card. That keeps it with the seat holding the budget lines (PR test(lint): the two lazy-deps contract files declare their whole-package-import budget #23888).vi.mockcensus goes before the externalisation.objectstack-fleet commentedon Oct 16, 2026
Claim: PM loop round 17 (
batch 2) · 2026-10-16T20:36Z ·packages/lint's vitest stops inlining the linked@objectstack/specdist (test.server.deps.externalwith a realpath pattern and a positive control), so the first test file to import a lint entry no longer pays the cold transform. One PR,Fixes #23900.Session:
session_01HjR8RUwTPQ3TgbfKihmxp9(the dispatching seat; the dev runs as a cloud session the seat post names)Account:
os-elon-musk(the seat's linked user asGET /useranswers it; this card's assignee)Branch:
claude/issue-23900-lint-vitest-external-specWorktree:
objectstack-issue-23900(in the cloud session's container)Domain:
domain:specSeat:
domain:spec#1(seat post #6017)File surface (paths on
origin/main2b274600b1or later; stop on a breach and explain it in the report):packages/lint/vitest.config.ts: onetest.server.deps.externalkey with its header comment, and the positive control the card asks for (a test or a config-time assertion that the pattern really matches the linked dist path).// test-budget:lines PR test(lint): the two lazy-deps contract files declare their whole-package-import budget #23888 declared onlazy-deps.test.tsandruntime-lazy-deps.test.ts, removed or re-measured under ruling6181973363item 2.Container & model:
S,mode:cloud(the maintainer's word in this session's chat: 「batch 调到 5, 可以开云卡」, now 「并发降到2」),model: default tier. Review: the seat's own record (test config only, no contract face).Clause-②: no
Responsibility: n/a — a test-time follow-up of test budget (domain:spec): 26 test file(s) over 10 s of wall in the hourly full battery (Part of #23151) #23180's lint slice (dev report
6214605271, open question 1)Thread-read: 6215866252
Ruling-ref: 6181973363 (the test-budget ruling; ci: merge-queue
Test Corereds on the new suite-duration ceiling —@objectstack/cloud-connectionmeasured 177.5 s and 200.3 s vs its 163.89 s ceiling, in two groups of PRs that do not touch it (live since1bd612bd6) #22949 ruling E item 3)Serial constraints cleared:
batch 2: this seat's other dev in flight is spec(record-surface):deriveRecordFlowSurface's shipped docblock closes the override list withouteditMode, which the same package declares and the renderer honours; the file header still says create/edit open a "full page" #23964 (claim6222996144, patch round). invitations: a delegated_admin's Invite User dialog offers the Owner and Admin roles the server then refuses — sys_invitation.role options are static, and no contract tells the client which roles the caller may grant #23957's PR fix(identity): Invite User offers only the roles the caller may grant, read from the role cap's own table #24033 is ACCEPTed and in the queue; its dev session is archived.packages/lint/vitest.config.tsor the two lazy-deps test files. No open claim names them.objectstack-fleet commentedon Oct 16, 2026
os-dev-report
{ "issue": 23900, "status": "done", "branch": "claude/issue-23900-lint-vitest-external-spec", "pr": "https://github.com/objectstack-ai/objectstack/pull/24048", "session": "session_019oKRRCxbLy3xARM3CH99sf", "premise_still_valid": true, "summary": "packages/lint/vitest.config.ts gains one test.server.deps.external entry, /[\\\\/]packages[\\\\/]spec[\\\\/]dist[\\\\/]/, matching the realpath of the linked spec dist, and its header now states both settings and why each is package-wide. New positive control src/vitest-spec-external.test.ts: it asserts that the spec root and /data modules the tests import are the very instance Node's loader holds (require(esm) through createRequire of the exports-named ESM entry, compared by export identity). It goes red at the identity assertions when the entry is deleted or replaced by the name-shaped /@objectstack[\\\\/]spec/ pattern. The mock census is 0 (vi.mock / vi.doMock / vi.hoisted) with a 30-file control leg in packages/cli. Package hold went from 84 s to 53-54 s, and the two lazy-deps files' single-file walls from 14.2-15.2 s to 6.5-6.8 s, so their two 30 s whole-package-import budget lines were dropped (not re-measured: no post-change CI reading exists). The claim on the card names this branch (comment 6224227197), and the card's assignee os-elon-musk was left untouched.", "tests": "At HEAD 77f016c5, pnpm --filter @objectstack/lint test under os-verify-lock.sh: 'Test Files 142 passed (142) · Tests 6804 passed (6804)', VERDICT command-exit 0, lock held 54s. Paired measurement on the same box under the lock: BEFORE at 2b274600b1, full package 141 files / 6802 tests green, hold 84 s (transform 23.74s, import 189.81s); lazy-deps.test.ts alone 15.01 s / 15.19 s, runtime-lazy-deps.test.ts alone 14.25 s / 14.63 s (each run exit 0). AFTER at be2f7c34, hold 53 s (transform 8.05s, import 96.39s), 142 / 6804 green; lazy-deps 6.77 s / 6.79 s, runtime-lazy-deps 6.79 s / 6.54 s (each exit 0). Ablation of the config (one-time, through scripts/ablation-replace.mjs on committed be2f7c34, anchor x1 to x0 and blob 4ecbe245f449 changed on disk; no build or dist involved, the config is read directly): (1) entry deleted, control 2/2 FAIL at lines 48 and 54 (toBe identity) after the typeof / toBeDefined lines pass; (2) name-shaped /@objectstack[\\\\/]spec/, the same 2/2 FAIL. Restore: disk blob equals the HEAD blob 4ecbe245f449, and git diff HEAD is empty. pnpm --filter @objectstack/lint typecheck: VERDICT command-exit 0, with check:test-typecheck OK (2 files / 6 errors held in the debt ledger, unchanged).", "gates": "Derived with node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack at 77f016c5 (4 paths, +85/-4): 51 families. Ran all 51 with the exit code captured before any pipe. 50 exited 0. pnpm check:dual-build-cjs-loads exited 3, PREREQUISITE NOT MET (needs 81 more packages built), so NOT MEASURED locally and left to CI. Three first-run exit-3s were cleared and re-run to exit 0: node scripts/check-plugin-teardown-shape.mjs --self-test (fetched its pinned fixture commit 621a4876), and pnpm check:lean-entry-closure (built the objectql closure). Extra, outside the derivation: node scripts/check-console-intercept-disarm.mjs exit 0 ('75 vitest-running package(s), every one disarms'). Reconciliation, node scripts/pm/dispatch-gates.mjs --repo objectstack-ai/objectstack --ran with a 'cmd :: exit N' record: '51 derived, 50 run, 1 NOT-MEASURED (1 DERIVED from a recorded exit 3), 0 UNRUN', exit 0. PR check-runs at report time on 77f016c5: success for Lint · PM dispatch-gates, Check PR Size, Part-of PR must not also close its card, The card this PR closes must claim this branch, No other open PR may claim the same issue, No other open PR may claim the same single-writer path, Check Documentation Links, Flag docs affected by code changes, Dogfood Regression Gate and Auto Label. in_progress for Test Core 1-6/6, Build Core, Lint · ESLint / repo & tooling / CI, release & engine / docs & code guards / the scoped lanes, the Type Check lanes, Governed Surface Queue Guard, and Check Changeset (a re-run after the label shows skipped). The CI verdict is not awaited here, per os-dev.", "mcp_calls": "3 — mcp__github__issue_read (get, #23900), mcp__github__issue_read (get_comments, #23900), mcp__github__pull_request_read (get_check_runs, #24048); all reads, no MCP write tool", "api_writes": "3 relay dispatches, each POST /repos/objectstack-ai/objectstack/dispatches executed by fleet-write as objectstack-fleet[bot]: (1) pr_create, i.e. POST /repos/objectstack-ai/objectstack/pulls (draft) + POST /repos/objectstack-ai/objectstack/issues/24048/assignees [os-elon-musk], read back byte-identical (5043/5043); (2) label-write labels_add, i.e. POST /repos/objectstack-ai/objectstack/issues/24048/labels [skip-changeset], read back matching; (3) this os-dev-report comment, i.e. POST /repos/objectstack-ai/objectstack/issues/23900/comments. Plus git push (not REST)", "open_questions": [ { "question": "The two lazy-deps files' 30 s test-budget lines: drop them or re-measure them?", "options": [ "A drop both: the files return to the 10 s FILE_BUDGET_SECONDS line", "B keep a re-measured line" ], "recommendation": "A, taken. Local single-file walls are now 6.5-6.8 s, and the lines' stated reason (vite's transform of the inlined spec dist) no longer holds. A budget is a CI reading x 1.25, and no post-change CI reading exists, so any number would be invented. The before ratio of CI to local (18.0/15.1, 19.2/14.4) projects about 8-9 s in CI, under 10 s. If an hourly reading lands over 10 s, the duration gate reports it (it never reds a required job), and that reading is the measurement a re-declaration under ruling 6181973363 item 2 needs. Axes: the contract stays tight and true, with no stale declaration hiding a regression up to 30 s, and nothing new is added." }, { "question": "The card prompt's measurement baseline said 98 s to 60 s; this box read 84 s to 53 s. Which one stands?", "options": [ "A both: different boxes and loads, same direction, ratio about 0.63 in both" ], "recommendation": "A. Under shared-box conditions the ratios agree (60/98 = 0.61, 53/84 = 0.63)." }, { "question": "Worktree: the primary checkout already held the branch, so it was detached to let ../objectstack-issue-23900 check the branch out. Any concern?", "options": [ "A none: the cloud container is single-session, and the primary checkout is left detached at 2b274600b1 with nothing uncommitted" ], "recommendation": "A." } ], "out_of_scope_findings": [ "carrier: none · noted, not filed: @objectstack/formula and @objectstack/sdui-parser, lint's other two workspace deps, are still inlined from their dist in packages/lint's vitest. Unmeasured; the card scoped spec only. In the PR's Acceptance notes." ], "line_budget": 89, "deviations": [ "The first gate sweep exceeded the 10-minute foreground cap and was moved to the background by the harness; it finished, and I waited for it in the foreground by polling its results file in a bounded loop (no sleep, no watcher left running).", "pnpm check:console-intercept-disarm does not exist as a root script (pnpm exit 254, script not found, not a gate result); I re-ran it by its real path, node scripts/check-console-intercept-disarm.mjs, exit 0.", "check:dual-build-cjs-loads is NOT MEASURED locally (exit 3, prerequisite: full-repo build); CI measures it.", "The harness-written commit trailer is 'Co-authored-by: Claude' (model-free, per AGENTS.md), so there is nothing to report there; the PR body's footer uses the session-URL form AGENTS.md prescribes instead of the harness's robot-emoji line." ], "files_changed": [ "packages/lint/vitest.config.ts", "packages/lint/src/vitest-spec-external.test.ts", "packages/lint/src/lazy-deps.test.ts", "packages/lint/src/runtime-lazy-deps.test.ts" ] }Generated by Claude Code
objectstack-fleet commentedon Oct 16, 2026
Patch round → PR #24048 ·
domain:specseat 1 · sessionsession_01HjR8RUwTPQ3TgbfKihmxp9· 2026-10-16T21:27Z · on the dev report6224943566(head77f016c52a), same claim6224227197, the dev session reused.What the patch fixes (read by the seat on the head and on its check-runs):
Lint · ESLintis red, and soLint & Repo Gates. The annotation is onpackages/lint/src/vitest-spec-external.test.ts:29:import * as specRoot from '@objectstack/spec'breaks the repo'sno-restricted-importsrule (eslint.config.mjsabout :987). A namespace import of the root entry is refused; use subpath or named imports. ImportdefineStackby name from the root (the control needs one root export, not the namespace). Make the./dataimport named too (ObjectSchema), so the file reads one way. The control's logic and its two ablation legs stay as they are: re-run both legs after the edit and paste the results.dispatch-gates.mjs --commandsdoes not list the ESLint lane, so the 51-family sweep could not catch this. Runnode --stack-size=4000 node_modules/eslint/bin/eslint.js --no-inline-configover the four changed files and record the exit code. Then re-run the package's tests under the lock.The report's open questions, the seat's calls:
6181973363item 2.Accepted as is, noted for the record: the control relies on Node's
require(esm), which is on by default from Node 22.12. CI runsnode-version: '22', the latest 22.x. On an older 22.x the control fails loudly (ERR_REQUIRE_ESM), not silently, so it needs no change.The seat's ACCEPT follows a green head.
Generated by Claude Code
objectstack-fleet commentedon Oct 16, 2026
os-dev-report
{ "issue": 23900, "round": "patch round on seat record 6225029161", "status": "done", "branch": "claude/issue-23900-lint-vitest-external-spec", "pr": "https://github.com/objectstack-ai/objectstack/pull/24048", "head": "e0d24934", "session": "session_019oKRRCxbLy3xARM3CH99sf", "premise_still_valid": true, "summary": "Patch item 1: vitest-spec-external.test.ts now imports defineStack from '@objectstack/spec' and ObjectSchema from '@objectstack/spec/data' by name. The root namespace import broke no-restricted-imports (the SUBPATH_NAMES block of eslint.config.mjs). The control's logic is unchanged, and both ablation legs were re-run on the new head: red at the Object.is identity assertions (:48:25, :54:26), restore proven by blob. Item 2: ESLint was run locally over the four changed files (exit 1 on 77f016c5 reproduces the CI annotation; exit 0 on e0d24934); the package tests and typecheck were re-run under the lock and the gates re-reconciled. Item 3: the replacement Verification section is in pr_body_now_false. One commit, e0d24934, on top of 77f016c5; no history rewrite.", "tests": "At e0d24934: pnpm --filter @objectstack/lint test under os-verify-lock.sh: 'Test Files 142 passed (142) · Tests 6804 passed (6804)', VERDICT command-exit 0, lock held 54s. The control alone: 'Tests 2 passed (2)', exit 0. Ablation through scripts/ablation-replace.mjs on committed e0d24934 (config only, no build or dist): leg 1 deletes the external entry ('ok mutation landed: anchor 1 to 0, blob 4ecbe245f449 to c5a3c8fe0067'), giving 'Tests 2 failed (2)', 'AssertionError: expected [Function defineStack] to be [Function defineStack] // Object.is equality' at src/vitest-spec-external.test.ts:48:25 and 'expected [Function] to be [Function] // Object.is equality' at :54:26, exit 1. Leg 2 is the name-shaped /@objectstack[\\\\/]spec/ ('ok mutation landed: anchor 1 to 0, blob 4ecbe245f449 to 25fa0f95fab6'), the same two failures at :48:25 and :54:26, exit 1. Restore after each leg: 'blob after restore 4ecbe245f449'; HEAD blob equals the disk blob 4ecbe245f449; git diff HEAD empty and git status --porcelain empty. pnpm --filter @objectstack/lint typecheck at e0d24934: VERDICT command-exit 0, check:test-typecheck OK (2 files / 6 errors, ledger unchanged).", "gates": "ESLint (not in the --commands list, per the seat record): node --stack-size=4000 node_modules/eslint/bin/eslint.js --no-inline-config over packages/lint/vitest.config.ts, packages/lint/src/vitest-spec-external.test.ts, packages/lint/src/lazy-deps.test.ts and packages/lint/src/runtime-lazy-deps.test.ts. Exit 1 on 77f016c5 ('29:8 error * import is invalid because ... are restricted ... no-restricted-imports', 1 problem); exit 0 on e0d24934, no output. dispatch-gates --commands at e0d24934: the same 51 families as the first round (diffed identical). All 51 were run with exit codes captured before any pipe; pnpm check:lean-entry-closure first exited 3 (objectql dist absent in the recreated worktree) and exited 0 after building the objectql closure. --ran reconciliation with the 'cmd :: exit N' record: 'Run reconciliation — 51 derived, 50 run, 1 NOT-MEASURED, 0 UNRUN'; the 1 is pnpm check:dual-build-cjs-loads, exit 3, PREREQUISITE NOT MET (full-repo build), measured by CI. Extra: node scripts/check-console-intercept-disarm.mjs exit 0. dispatch-gates noted the tree is 6 commits behind origin/main and that none of them touches what it derives from; no merge was made, since none is called for (no conflict, no main-caused red). PR check-runs on e0d24934 at report time: success for Lint · ESLint, Lint & Repo Gates, TypeScript Type Check, Build Core, Governed Surface Queue Guard, Test Core (1/6), (5/6), (6/6), all Type Check · lanes, all Lint · lanes, Dogfood Regression Gate, Check PR Size, Check Documentation Links, Flag docs affected by code changes, the three claim/Part-of guards and Auto Label; Check Changeset skipped (skip-changeset label). in_progress: Test Core (2/6), (3/6), (4/6).", "mcp_calls": "2 this round: mcp__github__issue_read (get_comments, #23900, to read seat record 6225029161) and mcp__github__pull_request_read (get_check_runs, #24048); both reads, no MCP write tool. Round 1's 3 reads are listed in report 6224943566.", "api_writes": "1 this round: the relay dispatch POST /repos/objectstack-ai/objectstack/dispatches executing POST /repos/objectstack-ai/objectstack/issues/23900/comments (this os-dev-report). Plus git push of e0d24934 (not REST). No pr_create, no label write, no claim, no PR body edit.", "pr_body_now_false": { "section": "## Verification", "replace_whole_section_with": "## Verification\n\n- Head `e0d24934` (patch round): the control now imports `defineStack` from `@objectstack/spec` and `ObjectSchema` from `@objectstack/spec/data` by name. The root namespace import broke the repo's `no-restricted-imports` rule for `@objectstack/spec`. The identity check is unchanged.\n- ESLint over the four changed files, `node --stack-size=4000 node_modules/eslint/bin/eslint.js --no-inline-config ...`: exit 1 on `77f016c5` (one `no-restricted-imports` error at `vitest-spec-external.test.ts:29:8`), exit 0 on `e0d24934`.\n- Ablation legs re-run on `e0d24934` through `scripts/ablation-replace.mjs`. With the `external` entry deleted, and with the name-shaped `/@objectstack[\\\\/]spec/`, the control is 2/2 red at `:48:25` and `:54:26` (`Object.is` identity). Restore: the disk blob equals the HEAD blob `4ecbe245f449`, and `git diff HEAD` is empty.\n- `pnpm --filter @objectstack/lint test` at `e0d24934`: 142/142 files, 6804/6804 tests (`VERDICT command-exit 0`, lock held 54 s).\n- `pnpm --filter @objectstack/lint typecheck` (including `check:test-typecheck`) at `e0d24934`: `VERDICT command-exit 0`; the debt ledger is unchanged (2 files / 6 errors).\n- `node scripts/pm/dispatch-gates.mjs --repo objectstack-ai/objectstack --ran ...` at `e0d24934`: 51 derived, 50 run with exit 0, 1 NOT-MEASURED (`pnpm check:dual-build-cjs-loads` exited 3 because it needs a full-repo build; CI measures it), 0 unrun. Also ran `node scripts/check-console-intercept-disarm.mjs`: exit 0.\n- No publish: `@objectstack/lint`'s `files[]` is `dist`, README, CHANGELOG. None of the changed text is in the built `dist/` (0 hits), while a shipped symbol (`validateStackExpressions`) has 8 hits as the positive control.\n" }, "open_questions": [], "out_of_scope_findings": [ "carrier: none · noted, not filed: dispatch-gates --commands does not derive the ESLint lane for a changed .ts file, so the 51-family sweep cannot catch a lint red; the seat record names it, and the gap was covered this round by running ESLint directly. Not a defect class a/b/c under os-dev rule 3; recorded only." ], "line_budget": 89, "deviations": [ "The worktree had been removed after round 1 and was recreated from origin's branch tip 77f016c5, so the lint closure (spec, formula, sdui-parser, lint) and the objectql closure were rebuilt under the lock before measuring.", "The gate sweep ran as a harness background command and was waited on in the foreground with a bounded until-loop on a sentinel file; nothing is left running." ], "files_changed": [ "packages/lint/vitest.config.ts", "packages/lint/src/vitest-spec-external.test.ts", "packages/lint/src/lazy-deps.test.ts", "packages/lint/src/runtime-lazy-deps.test.ts" ] }Generated by Claude Code
objectstack-fleet commentedon Oct 16, 2026
ACCEPT → PR #24048 comment 6225804072 ·
domain:specseat 1 · sessionsession_01HjR8RUwTPQ3TgbfKihmxp9· 2026-10-16T22:22Z · seat review only (test config and tests, no contract face), after a patch round (6225029161, an ESLintno-restricted-importsred in the new control test); the heade0d249340cis green and goes to the queue in this act.packages/lint's vitest now loads the linked@objectstack/specdist with Node instead of transforming it. A positive control goes red the moment the pattern stops matching. The package hold went from 84 s to 53 s, and the two lazy-deps files' 30 s budget lines are dropped because both files now run in under 7 s. The cloud dev session is archived at this ACCEPT.Generated by Claude Code
objectstack-fleet commentedon Oct 16, 2026
Landed · PR #24048 →
c91af8ed64(Fixes #23900,packages/lint's vitest loads the linked spec dist with Node) ·domain:specseat 1 ·os-elon-musk· sessionsession_01HjR8RUwTPQ3TgbfKihmxp9· 2026-10-16T22:52Zc91af8ed64origin/maingit merge-base --is-ancestor c91af8ed64 origin/main→ exit 0c91af8ed64as at the reviewed heade0d249340c(ACCEPT6225804072)gh-readonly-queue/main/pr-24048-*ref remainsFixes #23900is the body's only closing keyword, and the squash message carries no other; this card closed by that PR, as intended, and nothing else didFor whoever reads the next test-budget refresh list (#23180):
lazy-deps.test.tsandruntime-lazy-deps.test.tsno longer declare a 30 s budget. Both ran in 6.5-6.8 s alone after the change (before: 14.3-15.2 s), and the CI-to-local ratio from their old readings projects about 8-9 s in CI. If an hourly reading afterc91af8ed64puts either file over 10 s, that reading is the measurement a re-declaration needs under ruling6181973363item 2. Readings from beforec91af8ed64measure the old, inlined config and are not evidence either way.pm:dispatchedis cleared in this act. The cloud dev session was archived at the seat's ACCEPT.Generated by Claude Code