fix(create-objectstack): a project name that starts with the template namespace scaffolds instead of being refused

[objectstack-fleet]

Fixes #22911
Clause-②: no

What this changes

create-objectstack blank-probe exited 1 with "1 object name(s) still carry the template's 'blank_' prefix ... This is a bug in the scaffolder". The rewrite itself was right (blank_note became blank_probe_note). The self-check after it, findStaleNamespacePrefixes, tested each name against the template prefix alone, and blank_probe_note starts with blank_ because the new namespace blank_probe does.

Per the triage direction (comment 6115833632), a name is now stale only when it starts with the old prefix and does not start with the new namespace's prefix:

• packages/create-objectstack/src/rewrite-identity.ts: findStaleNamespacePrefixes(dir, oldNs, newNs) takes the new namespace as a required third argument. Each old-prefixed name: literal on a line is checked against the same namePrefixRe built from the new namespace. One predicate for every template namespace, with no exemption for blank. The check runs per literal, so a correct name on a line cannot hide a stale one on the same line.
• packages/create-objectstack/src/index.ts: the one caller passes the namespace it has just written. Nothing outside the package imports the function. The only other rewrite-identity mentions are comments, and the package's exports map publishes ./created-summary only.
• Error text and rewrite logic are unchanged.

Measured at the public door

Repo-built CLI (tsx packages/create-objectstack/src/index.ts blank-probe --skip-install --skip-skills):

tree	exit	result
base eab7236ff7	1	Scaffolding rewrote the namespace to 'blank_probe' but 1 object name(s) still carry the template's 'blank_' prefix: src/objects/note.object.ts:6 name: 'blank_probe_note',
this PR 8679418942	0	src/objects/note.object.ts:6 name: 'blank_probe_note', and objectstack.config.ts:27 namespace: 'blank_probe',

Tests

• rewrite-identity.test.ts: the two existing findStaleNamespacePrefixes cases now pass the new namespace. A new describe.each runs over blank → blank_probe and todo → todo_list, so the same pins run under two template namespaces. It has three cases:
  • after the rewrite, nothing is stale;
  • CONTROL: without the rewrite, the genuinely stale OLD_note is still reported;
  • a line carrying a correct name and a stale one is still reported.
• scaffold-namespace-extends-template.test.ts (new) drives the real CLI through tsx, the pattern scaffold-description.test.ts uses. There is one case per bundled template in TEMPLATES. Each probe project is named from that template's own namespace (blank gives blank-probe). The case asserts:
  • exit 0;
  • the scaffolded namespace is blank_probe;
  • every template-prefixed name moved to blank_probe_*;
  • each name passes the spec's own validateObjectNamespacePrefix.
• Runs at 8679418942, through the shared verify lock:
  • pnpm --filter create-objectstack exec vitest run --maxWorkers=2 followed by pnpm --filter create-objectstack run typecheck: Test Files 18 passed (18), Tests 261 passed (261), tsc --noEmit clean, VERDICT command-exit 0.
  • tsc --noEmit --listFiles includes both touched test files (18 test files in the program).

Ablation (committed fix, two legs, scripts/ablation-replace.mjs WRAP mode with a restore trap)

The anchor was if (literals.some((literal) => !newRe.test(literal))) {. For each leg the on-disk count went anchor 1 → 0 and replacement 0 → 1, and the blob changed.

• Leg A: if (literals.length > 0) {, the pre-fix predicate (old prefix alone). Predicted red; observed red: Tests 3 failed | 26 passed (29). The failures were both "reports nothing once the rewrite has run" cases and the scaffold case, which printed the original ✗ Scaffolding rewrote the namespace to 'blank_probe' but 1 object name(s) still carry the template's 'blank_' prefix.
• Leg B: if (literals.length > 0 && !newRe.test(lines[i])) {, a line-level predicate. Predicted red; observed red: Tests 2 failed | 27 passed (29). The failures were both "judges each literal" cases.
• Restore, after each leg and again in the trap: blob == HEAD (35eae3f4e754) and git diff HEAD is empty, porcelain empty.

No dist/ is involved: the unit tests import ./rewrite-identity.js from src/, and the scaffold test runs src/index.ts through tsx.

Edge answers (measured, one-off probe on 8679418942)

case	old → new	name on disk	stale
forward, rewritten	blank → blank_probe	blank_probe_note	0
forward, no rewrite	blank → blank_probe	blank_note	1
reverse, rewritten	blank_probe → blank	blank_note	0
reverse, no rewrite	blank_probe → blank	blank_probe_note	0
equal	blank → blank	blank_note	0
disjoint, no rewrite	blank → my_app	blank_note	1

The "reverse, no rewrite" row reads 0 because blank_probe_note is legal under namespace blank (blank_ + probe_note). The project would not be rejected, which is the question the check answers. The caller never calls the check for "equal" (namespace !== templateNamespace guard).

Gates (at 8679418942)

node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack, with no paths, derived 61 commands from merge base eab7236ff (5 paths, +153 / -8). That matches the PM's dispatch-time count. All 61 ran with the exit code captured before any pipe:

• 59 exit 0. Examples:
  • check-adr-0087-registration --base origin/main: "this PR adds no declared-breaking changeset (1 non-breaking changeset(s) seen)"
  • check:nul-bytes: "OK (scanned 10960 text file(s) ... no raw ASCII control bytes)"
  • check:cross-package-test-inputs: "OK: 30 package(s) read outside themselves, all declared"
  • check:test-source-alias: "OK — 73 packages with tests scanned"
  • check:issue-citations: "no issue citations added against eab7236"
  • check:published-files: "70 publishable package(s) ... declare a files whitelist"
  • check:type-check-debt: "--re-measure: OK — 1 ledger entr(ies) re-measured ... none above its recorded number"
• pnpm check:dual-build-cjs-loads, exit 3, NOT MEASURED: "PREREQUISITE NOT MET — this gate reads built output, and some package has no dist/ ... ⛔ This is NOT a pass: nothing was measured." It needs a whole-repo pnpm build. create-objectstack ships ESM only. CI's Build Core runs it.
• pnpm check:lean-entry-closure: exit 3 on the first pass ("PREREQUISITE NOT MET ... packages/objectql/dist/core.mjs"). After pnpm --filter '@objectstack/objectql...' build (under the lock, VERDICT command-exit 0), the re-run gave exit 0: "✓ check-lean-entry-closure: 2 published condition(s) measured from a real load".
• Reconciliation: dispatch-gates --repo objectstack-ai/objectstack --ran FILE, with each line recorded as COMMAND :: exit CODE, printed "61 derived famil(ies) accounted for — 60 run, 1 NOT-MEASURED (1 DERIVED from a recorded exit 3)".
• Also run: the four artifact-roster families whose roster sits under a directory this diff touches, plus the scaffold one. All exit 0:
  • check-changeset-fixed: "fixed group is in sync with 70 public workspace packages"
  • check:authz-resolver: "single shared authorization resolver intact"
  • check:error-code-casing: "no unlisted lowercase error codes in 8216 scanned file(s)"
  • check:filter-alias-parity: "OK (4 transport spelling(s) ...)"
  • check:scaffold-emission-policy: "2 carrier(s) emit the shared policy"

Lint (proven narrowing; pnpm lint is CI's)

• Population (from eslint.config.mjs): files: ['**/*.{ts,tsx,mts,cts,js,jsx,mjs,cjs}'] minus NEVER_LINTED. eslint --print-config returns a config for each of the 4 touched .ts files (5 to 6 rules each), so none is ignored.
• Count (from --format json): 4 files, 0 errors, 0 warnings, run with --no-inline-config as pnpm lint does.
• Invariance: the config never enables type-aware linting. Every parserOptions is { ecmaVersion, sourceType }, with no project or projectService. Its only disk reads are its own baselines, so this diff cannot move a verdict on any untouched file.

Acceptance notes

• The "second template" pin does not apply today. TEMPLATES in template-registry.ts holds only blank, and src/templates/ holds only blank/ and AGENTS.md. The scaffold-level pin iterates TEMPLATES and names each probe from that template's own namespace, so a second template is pinned on registration.
• The reverse case (a template namespace X_Y with project namespace X) cannot be reached with the bundled template, because blank contains no underscore. If a rewrite were skipped there, the check would pass names that are already legal under the new namespace. That is consistent with the decided predicate, not a defect.
• origin/main moved two commits after this branch was cut (3caae97942, 018b077fbf). Both touch packages/objectql only, and git diff --stat eab7236ff7 origin/main -- packages/create-objectstack is empty. They were not merged in; the merge queue rebuilds on main.
• Changeset .changeset/22911-blank-prefixed-project-name.md is create-objectstack: patch, with Clause-②: no.

---

Generated by Claude Code

[github-actions]

📓 Docs Drift Check

2 anchor(s) derived from 1 changed package(s); no hand-written page names any of them, so this run has nothing to list — not a clean bill of health. This check sees only pages that NAME a derived anchor: one that documents this change in prose, or enumerates it in an authoring dialect, names none and stays invisible to it on every run.

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 — 11 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 018b077fbf8876c0fd5f46e66dcc551a4e74c895 → packageMentionDocs.

Which tree this was computed on

This run read content/docs from 9c9f68e6feb41c1cdcc4176c9c227553ebd0f572 — the merge of head 8679418942c91827751803a3eda0ff92c724165b into base 018b077fbf8876c0fd5f46e66dcc551a4e74c895, 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 9c9f68e6feb41c1cdcc4176c9c227553ebd0f572 && git checkout 9c9f68e6feb41c1cdcc4176c9c227553ebd0f572
# afterwards, rebuild it from the two parents, which stay fetchable
git fetch origin 018b077fbf8876c0fd5f46e66dcc551a4e74c895 8679418942c91827751803a3eda0ff92c724165b && git checkout -B drift-repro 018b077fbf8876c0fd5f46e66dcc551a4e74c895 && git merge --no-ff 8679418942c91827751803a3eda0ff92c724165b

node scripts/docs-audit/affected-docs.mjs --json 018b077fbf8876c0fd5f46e66dcc551a4e74c895

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