fix(lint): a liveness finding's fix text never comes from the ledger note, for any row class

[objectstack-fleet]

Fixes #21096

Clause-②: no

What

checkItem in packages/lint/src/lint-liveness-properties.ts now builds the fix text as entry.authorHint ?? defaultHint. The row's ledger note is never author-facing, for every row class (opted-in authorWarn, experimental, dead, live-elsewhere, planned). The now-unused isVerdictTriggered helper and the comments describing the old split are removed. describe() already carries a default hint for all four verdicts a finding can carry (experimental, planned, dead, live-elsewhere), so no prose was invented. A live row with authorWarn still throws the integrity sentinel, unchanged.

No ledger row under packages/spec/liveness/** is edited: this is the structural direction from triage, not per-row authorHint discipline.

Measured end to end

os lint --json on examples/app-showcase (built from this tree, before and after rebuilding @objectstack/lint), the two liveness-planned-property findings, fix field:

• Before, both findings: "[ADR-0090 D11] P1 lands the SPEC SHAPE only (validated at authoring that the external dial is no wider than the internal one; ...) ... tracked on ADR-0090: Permission Model v2 — concept convergence, final naming, AI-authoring safety (tracking) #2696. planned + authorWarn per enforce-or-mark ... Re-verified 2026-07-30: ..." (about 700 characters of maintainer prose, one tracker id).
• After, both findings: "Keep it — a consumer is being built against this property; it has no runtime effect yet."

The two liveness-dead-property findings on showcase_contributor are unchanged ("Remove it — it is declared in the spec but not consumed at runtime."). Exit code 0 before and after.

Tests (lint-liveness-properties.test.ts)

• The real-ledger object.externalSharingModel pin now asserts the planned default hint, no # followed by digits, and that the row's note text is absent (with an anti-vacuity guard that the row is still opted-in, has no authorHint and a note with a tracker id).
• New real-ledger pin over every shipped opted-in and experimental row: a row with an authorHint prints it exactly (control); a row without one prints neither its note nor a tracker id.
• New synthetic pin: authorHint wins, otherwise the verdict default, never the note, for dead, live-elsewhere, planned and experimental, with and without authorWarn.
• Two feat(lint): a ledger-dead or live-elsewhere key warns without an authorWarn opt-in, and never shows the ledger note (#16094) #21092 pins that encoded the old "opted-in rows show the note" behavior are rewritten.

Results: pnpm --filter @objectstack/lint test 118 files / 5486 tests pass; the liveness file alone 93 pass; pnpm --filter @objectstack/lint typecheck green.

Ablation (one-line revert to entry.authorHint ?? entry.note ?? defaultHint, through scripts/ablation-replace.mjs, anchor hit 1 time, blob moved, restore verified byte-identical to HEAD): 6 tests go red (the two real-ledger pins, the two synthetic pins, the live-elsewhere precedence pin, the dead RLS end-to-end pin); 87 stay green. Restored; git diff HEAD empty.

Other sites where a ledger note could reach an author

None found in packages/lint/src or packages/cli/src besides this expression. The other .note reads in validate-react-page-props.ts are deprecation records, not ledger rows. Nothing else reads the ledger's note as author text.

Gates

node scripts/pm/dispatch-gates.mjs --commands derived 60 families; each was run and recorded with its exit code, then reconciled with --ran: 59 exit 0, 1 NOT MEASURED (check:dual-build-cjs-loads, exit 3, PREREQUISITE NOT MET: eight unrelated packages have no dist/ in this worktree; nothing was measured, and the diff touches no exports or build output). CI owns the rest.

Coordination

#21135 (the liveness README) states the same model. It has no open PR and no branch at this time; whichever lands second keeps the two consistent. This PR does not touch the README.

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

This PR changes 1 package(s): @objectstack/lint, touching 4 documentable anchor(s).

⛔ 2 release-owned page(s) name something this change touched. These are read-only:

• content/docs/releases/v14.mdx (via authorWarn (literal, a string literal in a comment on a changed line))
• content/docs/releases/v17/index.mdx (via authorWarn (literal, a string literal in a comment on a changed line))

content/docs/releases/ is RELEASE-OWNED (AGENTS.md "Documentation Guardrails"): release
notes are written centrally at release time, and a code PR that edits them is the exact PR
that guardrail exists to stop. They are still audited — read-only. If one of them is actually
wrong, file an issue or open a dedicated docs-only PR; do not edit it here.

What this run could not see

• the SDK route bridge reached 54 of 206 client-bound route-ledger rows — the other 152 have no registrar path: tail to select them, so pages documenting THEIR client methods cannot appear above, on this or any run. Of those 152: 0 are remediable by widening that discovery convention (an in-repo file declares the path; the convention did not scan it); 55 are structural — on a ledger where NOT ONE row is declared in-repo, so no discovery change reaches them at any price; 97 are undecided (no in-repo declaration, on a ledger that has other in-repo registrars — absence and an unreadable spelling are not distinguishable here). The rows themselves: node scripts/docs-audit/affected-docs.mjs --bridge-coverage
• a page that states a rule by its inputs shares no identifier with the emitter that implements the rule, so an emitter-only diff cannot list it — not on this run and not on any run. Measured on fix(driver-sql): emit varchar(maxLength) for a text field a declared index keys on #11430: content/docs/protocol/objectql/types.mdx documents the text-family column mapping by the ObjectQL type names it maps FROM (text / textarea / html) while the diff changed createColumn; it went unlisted, and it was the page that diff falsified, in four places. No shared token exists to detect this on, so a rule your change carries has to be re-read by hand in the pages that restate it.
• a key NAME is not a key, so the hand re-read the line above prescribes can land on the wrong schema. The same spelling is authorable on one governed type and a [REMOVED] tombstone on another for each of active, aria, joins, objects, template, tools and version (censused on [finding] tools is a key on BOTH AgentSchema (tombstoned, dead) and SkillSchema (live, cloud-attested), so a name-based search attributes skill examples to the agent key — it produced a false stop-the-line alarm on PR #19059 #19093 over the liveness ledger's governed types, top-level keys); nothing in a search result distinguishes the two, so a grep hit on a LIVE example reads as evidence about the DEAD key. Measured on fix(spec): the agent.tools liveness row says dead — it claimed live on a key the schema tombstoned #19059: content/docs/ai/agents.mdx was reported as contradicting the agent.tools tombstone over its tools: example at :161, which is inside the defineSkill({ block opened at :155 — the page was already correct. Settle ownership by PARSING the value against both schemas, never by the name: that literal PASSES SkillSchema, and as an AgentSchema it FAILS at tools with the tombstone prescription. ⛔ These names are not the whole class — a key retired through a .strict() guidance map leaves no tombstone in the walked shape and none of them here (tool.category, live as AIToolDefinition.category).

Coarse fallback — 4 page(s) merely mention a changed package (the pre-#9192 predicate, kept for the deliberately-wide backstop): node scripts/docs-audit/affected-docs.mjs --json f0cc16e8d55b02269c7db708254e2917ba7e84e2 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from 3a8e95838807d9a98c9d997f2b631d9add3ffc3c — the merge of head 88dc06a2012516079c9d5b70230ce23a2ee18df5 into base f0cc16e8d55b02269c7db708254e2917ba7e84e2, which is what actions/checkout gives a pull_request run. Not the PR head.

A worktree cut from an older main holds a different content/docs, so re-deriving there can legitimately return a different list — that is a different tree, not a wrong row. To answer on the same tree:

# while this PR is open — GitHub drops the merge commit once it closes
git fetch origin 3a8e95838807d9a98c9d997f2b631d9add3ffc3c && git checkout 3a8e95838807d9a98c9d997f2b631d9add3ffc3c
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin f0cc16e8d55b02269c7db708254e2917ba7e84e2 88dc06a2012516079c9d5b70230ce23a2ee18df5 && git checkout -B drift-repro f0cc16e8d55b02269c7db708254e2917ba7e84e2 && git merge --no-ff 88dc06a2012516079c9d5b70230ce23a2ee18df5

node scripts/docs-audit/affected-docs.mjs --json f0cc16e8d55b02269c7db708254e2917ba7e84e2

⚠️ That checkout carried uncommitted changes, so the commit above does not fully identify what was read.

Advisory only, and a precision-first one (#9192): a page is listed because it names a
symbol, wire route or SDK method this diff touched — not because it mentions a changed
package. Each row says which anchor put it there, so a wrong row is reportable rather than
merely annoying. To re-verify, run the docs-accuracy-audit workflow scoped to these files:
node scripts/docs-audit/affected-docs.mjs f0cc16e8d55b02269c7db708254e2917ba7e84e2 → pass the list as
args.docs, on the commit named under Which tree this was computed on.

[objectstack-fleet]

Contract review

Served-tier: CONTRACT_REVIEW_TIER
Head-sha: 88dc06a2012516079c9d5b70230ce23a2ee18df5
Local-runs: none

① Derived judgments

The checkout has no node_modules, so the one permitted vitest file was not run. Every judgment below comes from reading the diff, the branch source, the main-side test and the shipped ledger JSON.

Code (packages/lint/src/lint-liveness-properties.ts).

• const hint = entry.authorHint ?? defaultHint; is correct for every class that can reach it.
  • describe(entry) returns a non-empty defaultHint for experimental, planned, dead and live-elsewhere.
  • Any other status (live + authorWarn: true, or a missing status + authorWarn: true) throws the integrity sentinel inside describe() before hint is computed.
  • The shipped ledgers have no live + authorWarn row at the depth the warn map reads (720 live, all unwarned), so the sentinel is not newly reachable.
• Nothing is left dangling.
  • isVerdictTriggered has zero remaining references.
  • RULED_VERDICTS is still consumed by VERDICTS_THAT_WARN.
  • The remaining note mentions are the LedgerEntry.note field, the lintLivenessProperties tells the author a planned property is dead — the message and the hint on the same finding contradict each other #11384 history paragraph, and the two rewritten comments. The header comment, the describe() doc and the inline comment all state the code truthfully.

Tests (packages/lint/src/lint-liveness-properties.test.ts).

• The object.externalSharingModel real-ledger pin guards that the row is still authorWarn: true, has no authorHint, and carries a note matching /#\d+/; all three hold, and the note cites ADR-0090: Permission Model v2 — concept convergence, final naming, AI-authoring safety (tracking) #2696. It then asserts the planned default hint, no tracker id, and not the note. This is strictly stronger than main's not.toMatch(/^Remove it/), and it would fail on main.
• The opted-in/experimental real-ledger pin.
  • Its three anti-vacuity guards are non-vacuous against the shipped ledgers: 1 opted-in row and 5 experimental rows have a note and no authorHint, and 1 of those notes carries a tracker id.
  • The authorHint rows (3) are pinned byte for byte. Main's CONTROL (authorHint ?? note) is correctly inverted, not weakened.
  • For rows without an authorHint, this pin asserts only "not the note, no tracker id". The synthetic every-class pin and the externalSharingModel pin assert the default.
• The synthetic every-class pin covers dead, live-elsewhere, planned and experimental, each with and without authorWarn, and asserts the default. A control asserts that authorHint wins. It keeps every main-side assertion for verdict-triggered rows and inverts exactly the opted-in/experimental ones.
• The feat(lint): a ledger-dead or live-elsewhere key warns without an authorWarn opt-in, and never shows the ledger note (#16094) #21092 pins. The verdict-triggered real-ledger pin and the end-to-end dead-RLS pin are untouched. The live-elsewhere precedence pin moves from toBe('N') to toContain('sibling repo'), which is the required inversion. No pin was dropped without a stronger replacement.

Changeset (.changeset/21096-liveness-hint-never-note.md).

• "728-character maintainer note that cites a tracker id": TRUE. packages/spec/liveness/object.json → props.externalSharingModel.note is 728 characters and contains #2696.
• "the two liveness-planned-property findings for externalSharingModel" on app-showcase: TRUE (announcement.object.ts:34, account.object.ts:26).
• The quoted fix text: TRUE, byte-identical to the planned defaultHint in describe().
• "an opted-in row and an experimental row with no authorHint printed its note": TRUE of the main-side expression.
• "No rule id, message, severity or exit code changes": TRUE; only the hint expression changed.
• "a row that has an authorHint prints it exactly as before": TRUE, pinned by the CONTROL loops.
• The prose names no tracker id other than (#21096): TRUE.

Merge-cleanliness. git merge-tree --write-tree origin/main FETCH_HEAD exits 0 (tree 7f3621751107ec08904c113c2671097e86aee959), with no conflicts.

② Semver level

patch for @objectstack/lint is right: a bug fix in a released package, with no authorable surface or rule id changed. Clause-②: no is right.

③ Boundary flags

• A stale maintainer doc outside the three files. packages/spec/liveness/README.md:558 still says "authorHint" "(falls back to note)". It was already partly wrong after feat(lint): a ledger-dead or live-elsewhere key warns without an authorWarn opt-in, and never shows the ledger note (#16094) #21092, and it is now wrong for every row class.
  • Not blocking: it is a ledger-maintainer README, not an author-facing or generated surface. The PR body scopes itself to the lint package and hands the README to [finding] packages/spec/liveness/README.md "Author warnings" still says warnings are opt-in per authorWarn row; since #21092 a dead or live-elsewhere verdict warns on its own #21135 (queued), which states the same model.
• No other reader of the note-fallback precedence exists in src, docs or scripts.

Implemented-by: claude/issue-21096-liveness-hint-no-note
Reviewed-by: session_01VDtqoecgES7ScQYGbFVDRv

VERDICT: PASS

---

Generated by Claude Code