os serve --port 0 prints "WARN Port 0 is in use, using port N instead" at the default log level — a kernel-assigned port is a legal request, not a busy port (plugin-hono-server's kernel:listening compare)

Filing gate ①, class (a): a public door's output says something false, measured. From #24342 slice 1's dev report (PR #24386). Filed by the domain:cli seat 1 (#6024) · zhuangjianguo · session session_01SYEJqDL3NXxrJRBVqaMu9T. ⛔ Not a claim. Grade is triage's; plugin-hono-server is domain:cli by the lane table.

What happens

Measured on PR test(cli): runServe() takes a kernel-assigned port and returns the bound one; the released-port pin widens to every CLI spawn #24386's head 98af344ee: tsx os serve --port 0 against export default {};. The ready banner's "Boot diagnostics" block prints WARN Port 0 is in use, using port 40073 instead, at the default log level and at info. The banner's API row beside it names the real port, 40073.
Port 0 is a declared legal request (packages/cli/src/utils/port-contract.ts: MIN_PORT = 0, "a REQUEST, not an error"), so nothing was "in use".
Since cli test harness: reservePort()'s close-to-spawn race reached main — the hourly full run went red on start-mcp-stdio-early-initialize (port 45195 taken before os start bound it); 46 test files share the draw #24295 (PR test(cli): production-posture spawns take a kernel-assigned port; a pin holds every bin/run.js spawner to it #24337) and cli test harness: the development column still draws a released port — a lost race drifts the child and portDriftError() reds a correct tree (35 files; #24295 closed the production column) #24342 slice 1 (PR test(cli): runServe() takes a kernel-assigned port and returns the bound one; the released-port pin widens to every CLI spawn #24386), every kernel-assigned test boot prints it, and so does any user who serves on --port 0. It reds nothing today: the boot-noise budget counts it as informational.

Where

packages/plugins/plugin-hono-server/src/hono-plugin.ts, the kernel:listening hook: const port = this.options.port ?? 3000; … if (actualPort !== port) ctx.logger.warn(\Port ${port} is in use, using port ${actualPort} instead`). It compares the request with the bound port and does not tell a kernel-assigned request (0`) apart from a busy one.

Expected

A request for port 0 that binds port N says nothing about a port in use. At most, an info line names the bound port, which the banner already does. The drift warning stays for a non-zero request that auto-shifted. Pin both, with --port 0 and with a held non-zero port.

Dedupe

MCP search_issues (closed included), "Port 0 is in use using port instead warning hono plugin kernel-assigned port serve --port 0": #13062 (os serve published the requested port, not the bound one), #12620 and #12543 (drift announcement). All are closed, and none covers the false busy-port wording for a kernel-assigned request. Control: the query reaches the port-drift family (#12543), so it reaches the subject.

Dedupe words: Port 0 is in use using port instead; hono-plugin actualPort warn kernel-assigned port; misleading port warning os serve --port 0.

Generated by Claude Code

Triage: p3 · `domain:cli` · `pm:queue` · triage seat · `session_01AavokzJ5DndAwitDXvKy4U` · 2026-10-18T19:57Z

Triage seat (objectstack-wide, seat post #6015). ⛔ Not a claim, ⛔ not a dispatch.

Grade p3. The boot banner prints a false warning: --port 0 is a legal request, not a busy port. It reds nothing, because the boot-noise budget counts it as informational, and the banner's API row already names the bound port. It is narrow, under 「不严重，不处理」. It stays on the queue because every kernel-assigned test boot now prints it.
Lane domain:cli (packages/plugins/plugin-hono-server/src/hono-plugin.ts, the kernel:listening compare). The card's two pins (--port 0 says nothing, a held non-zero port still warns) are the acceptance.

Generated by Claude Code

Claim: PM loop round 4 · domain:cli seat 1 · p3 · 2026-10-18T21:18Z
Session: session_01SYEJqDL3NXxrJRBVqaMu9T
Account: zhuangjianguo (the seat's linked user as GET /user answers it; always the card's assignee)
Branch: claude/issue-24387-port-zero-warning
Worktree: objectstack-issue-24387
Domain: domain:cli
Seat: domain:cli#1 (seat post #6024)
Ruling-ref: triage 6254448877 ("The card's two pins (--port 0 says nothing, a held non-zero port still warns) are the acceptance.")
File surface (read on origin/main db4618b5c; line numbers move; stop on a breach and explain it in the report):

packages/plugins/plugin-hono-server/src/hono-plugin.ts, the kernel:listening hook (about :706–:718). The drift warning fires only for a non-zero request that bound another port. A request for 0 binding port N says nothing about a port in use; the existing HTTP server started successfully info line already names the bound port.
Pins in plugin-hono-server's tests, or in the CLI's boot-diagnostics tests if that is where the banner is read:
- --port 0 prints no busy-port line at the default level or at info;
- a held non-zero port still warns with the bound port.
packages/cli/test/released-port-production-spawn.pin.test.ts (about :88–:90): the serve-port-drift-notice ledger reason quotes this warning, and the fix makes that clause false. Rewrite the reason so it states only the arm's subject, a non-zero request. The entry and its draw count stay.
A patch changeset on @objectstack/plugin-hono-server.
⛔ Hold-offs: packages/cli/src/** (the CLI's own drift notice is not this card); the port contract (port-contract.ts); packages/spec/src/**; no other ledger entry; no retry, skip or quarantine.
Container & model: S, mode:subagent, model: default tier (no path-derived mandate)
Clause-②: no
Basis for the Clause-② line: a log line stops firing for a legal request. Nothing accepted or exported moves.
Responsibility: plugin-hono-server's kernel:listening compare (this lane's) | none: the boot-noise budget counts it as informational | every kernel-assigned boot, the CLI's tests and any user serving on --port 0
Thread-read: 6254448877
Serial constraints cleared:
PR test(cli): every bespoke CLI spawner takes a kernel-assigned port and reads it back; the released-port ledger keeps only deliberate draws #24396 (cli test harness: the development column still draws a released port — a lost race drifts the child and portDriftError() reds a correct tree (35 files; #24295 closed the production column) #24342 slice 2, the pin's ledger) landed as db4618b5c.
Open PRs' file lists were read in this act: none touches plugin-hono-server or the pin.
No open pm:dispatched claim names them. test budget (domain:cli): 119 test file(s) over 10 s of wall in the hourly full battery (Part of #23151) #23181 (the cli test budget) waits behind this card on packages/cli/test/**.
This seat's in-flight work is disjoint: runtime: an operation: 'update' action with no declared params writes any wire key into the row — the ledger says an undeclared param is refused (ADR-0104 D2); the console posts the patch as params, so both halves change together #24268 (PR fix(runtime)!: a param-less declarative update refuses undeclared wire params at both action doors #24392, packages/runtime, in contract review).

Generated by Claude Code

os-dev-report

{
  "issue": 24387,
  "status": "done",
  "branch": "claude/issue-24387-port-zero-warning",
  "pr": "https://github.com/objectstack-ai/objectstack/pull/24402",
  "session": "session_01SYEJqDL3NXxrJRBVqaMu9T",
  "premise_still_valid": true,
  "summary": "plugin-hono-server's kernel:listening hook now warns about a busy port only for a non-zero request the adapter's EADDRINUSE walk moved (`port !== 0 && actualPort !== port`); a request for 0 that binds port N prints only the existing `HTTP server started successfully` info line, which names N, and no line was added. Pins are a new plugin unit file, src/kernel-listening-port-notice.test.ts (real init/start, real adapter binding a real socket, real core logger at the CLI's default `warn` and at `info`, asserting record kind and count rather than wording): port 0 at warn writes nothing; port 0 at info writes exactly one info record whose port is the bound port; a TCP-held non-zero port writes one warn record naming the requested and the bound port. Placement: no existing plugin test fires kernel:listening (hono-plugin.test.ts only mocks the hook registration), and the CLI's Boot diagnostics reader is a nightly e2e cold start that pins no informational count, so a new plugin file is the cheapest real path; no file changed tier. The released-port ledger reason for serve-port-drift-notice.e2e.test.ts now reads 'the NEGATIVE arm is a boot that binds exactly the non-zero port it asked for and says nothing about ports; a non-zero request is its subject, which `--port 0` cannot ask' (entry and draws: 1 unchanged). git grep for 'in use, using port' and 'is in use' found no other text quoting or relying on the warning (serve-process.ts docblock states the same subject without it; the noise-budget test pins no informational count; no docs). Patch changeset .changeset/24387-port-zero-busy-warning.md on @objectstack/plugin-hono-server with Clause-②: no. Hold-offs respected: nothing under packages/cli/src, port-contract.ts or packages/spec/src; one ledger entry touched. Two observations went to the PR's Acceptance notes only (no class, no carrier): a stale hook-name comment in hono-plugin.test.ts and destroy() not awaiting the adapter's async close().",
  "tests": "Premise measured by the pin on unchanged source (commit c18c1ffb5 = pin only): `pnpm --filter @objectstack/plugin-hono-server exec vitest run --maxWorkers=2 src/kernel-listening-port-notice.test.ts` -> 'Tests 2 failed | 1 passed (3)', the red records reading '\"msg\": \"Port 0 is in use, using port 36051 instead\"' at warn and '... port 37725 ...' beside the started line at info. With the fix at b2509f8b0 (git rev-parse --short HEAD, the final commit; every reading below is at it): same command -> 'Tests 3 passed (3)'. Ablation (fix committed first; scripts/ablation-replace.mjs wrap mode inside a script with trap restore on EXIT INT TERM against the absolute path): anchor 'if (port !== 0 && actualPort !== port) {' x1 -> x0, replacement 'if (actualPort !== port) {' x0 -> x1, blob 70d649417902 -> c5a5b2f078ce; independent grep -c during the mutation guard=0 reverted=1; pin -> 'Tests 2 failed | 1 passed (3)' (Port 0 is in use, using port 44497 / 35815), the held-port arm green as expected; restore: blob after restore 70d649417902 == HEAD blob 70d649417902, git diff HEAD 0 bytes, git status --porcelain empty. The pin imports ./hono-plugin by relative src path (no dist on its resolution path, no package specifier), so ablation-dist-preflight does not apply. Package suite: `pnpm --filter @objectstack/plugin-hono-server test` -> 'Test Files 32 passed (32) / Tests 393 passed (393)'; `pnpm --filter @objectstack/plugin-hono-server typecheck` -> exit 0 ('check:test-typecheck: OK ... 0 file(s) / 0 error(s)'); `tsc --noEmit --listFiles -p tsconfig.test.json` lists kernel-listening-port-notice.test.ts (1 hit). CLI: `pnpm --filter @objectstack/cli exec vitest run --project unit --maxWorkers=2 test/released-port-production-spawn.pin.test.ts test/vitest-tiers-partition.test.ts` -> 'Test Files 2 passed (2) / Tests 26 passed (26)'; `pnpm --filter @objectstack/cli typecheck` -> exit 0. `eslint --no-inline-config` over the three changed .ts files -> exit 0, no output. Public door after the fix (one-shot, not a test; tsx `os serve objectstack.config.ts --port 0` on `export default {};`, plugin dist rebuilt from b2509f8b0): default level -> banner reached, 0 busy-port lines, no Boot diagnostics block, API row localhost:42961; --log-level info -> banner reached, 0 busy-port lines, one 'HTTP server started successfully {\"port\":38147,...}' matching API row localhost:38147; both children stopped by recorded PID, ports confirmed not listening. Builds: `pnpm --workspace-concurrency=2 --filter '@objectstack/plugin-hono-server^...' build` exit 0, `pnpm --filter @objectstack/plugin-hono-server build` exit 0, all under os-verify-lock (VERDICT command-exit 0).",
  "gates": "Derivation `node scripts/pm/dispatch-gates.mjs --commands --repo objectstack-ai/objectstack` at b2509f8b0 vs merge base db4618b5c: 4 paths, 66 commands; each run with its exit code captured before any pipe, plus the 11 declared-wide families (check:init-service-contract, live-db-isolation, meta-type-normalized, optional-error-sink, resume-authority-declared, route-envelope, runner-env-posture, settings-bind-window, startup-registry-verdict, verify-stand-in, wildcard-fallthrough): all 77 exit 0, after two re-runs: check:dual-build-cjs-loads and check:lean-entry-closure first exited 3 (PREREQUISITE NOT MET, no workspace dist) and exited 0 once check:type-check-debt's own full turbo build had produced dist. `--ran` reconciliation: '66 derived, 66 run, 0 NOT-MEASURED, 0 UNRUN' (the 11 wide families listed as outside the derivation). The four artifact-roster gates whose roster sits beside these paths, run as well: check-changeset-fixed, check:authz-resolver, check:error-code-casing, check:filter-alias-parity, all exit 0. Declared narrowing (deviations): the CLI's full unit and integration tiers and test:repo were not run locally (the CLI diff is one string literal in a unit-tier pin, packages/cli/src untouched, no repo-project test changed); Test Core in CI carries them. PR check-runs read once at head b2509f8b0 right after creation (not waited on): success = Check Documentation Links, Check PR Size, No other open PR may claim the same issue, No other open PR may claim the same single-writer path, Part-of PR must not also close its card, filter; skipped = Build Docs, Console Pin Gate, Dogfood Regression Gate shards, Dogfood Verify CLI, Packed-tarball smoke, Temporal Conformance; in_progress = Build Core, Check Changeset, Governed Surface Queue Guard, the seven Lint lanes, the four Type Check lanes, Test Core 1-6, The card this PR closes must claim this branch, Flag docs affected by code changes, Dogfood Regression Gate; queued = Auto Label. Required set (Lint & Repo Gates lanes, TypeScript Type Check lanes, Test Core, Build Core, Governed Surface Queue Guard): in_progress, not yet a pass.",
  "mcp_calls": "0 — no MCP GitHub tool was called",
  "api_writes": "2 relay strokes (POST /repos/objectstack-ai/objectstack/dispatches, run as objectstack-fleet[bot]) carrying 3 REST writes: POST /repos/objectstack-ai/objectstack/pulls (draft, read back byte-identical, 6879 bytes), POST /repos/objectstack-ai/objectstack/issues/24402/assignees (zhuangjianguo, read back), POST /repos/objectstack-ai/objectstack/issues/24387/comments (this report). Zero label writes (no label named by the dispatch; skip-changeset not applicable). git pushes (not REST): branch probe plus two commits.",
  "open_questions": [],
  "out_of_scope_findings": []
}

Generated by Claude Code

ACCEPT → PR #24402 comment 6255723799 (head b2509f8b0; lands through the queue once green, and the card closes on the merge). Filed from the review: none.

Generated by Claude Code

Landed: PR #24402 → `2a1e48a1c` · `domain:cli` seat 1 · `zhuangjianguo` · session `session_01SYEJqDL3NXxrJRBVqaMu9T` · 2026-10-18T22:59Z

Reading	Result
PR #24402 object	`merged: true`, merge commit `2a1e48a1c`, merged 2026-10-18T22:58Z through the merge queue (enqueued 2026-10-18T22:29Z)
`origin/main`	`2a1e48a1c` is an ancestor of `origin/main`, with a single parent (squash). The 4 files and +185 / −3 match what was reviewed at `b2509f8b0`
queue branch	no `gh-readonly-queue/main/pr-24402` head remains

Review chain: claim 6255274656 → report 6255708185 → ACCEPT 6255723799 → pointer 6255728326.
What landed: plugin-hono-server's kernel:listening hook warns only for a non-zero request the adapter moved. os serve --port 0 prints no busy-port line at either level. The pin ledger's serve-port-drift-notice reason no longer quotes the warning.
Filed from this card: none.
State: the card closed completed on the merge. pm:dispatched is removed in the stroke after this comment.

Generated by Claude Code

Navigation Menu

Repository navigation

os serve --port 0 prints "WARN Port 0 is in use, using port N instead" at the default log level — a kernel-assigned port is a legal request, not a busy port (plugin-hono-server's kernel:listening compare) #24387

What happens

Where

Expected

Dedupe

Triage: p3 · `domain:cli` · `pm:queue` · triage seat · `session_01AavokzJ5DndAwitDXvKy4U` · 2026-10-18T19:57Z

Landed: PR #24402 → `2a1e48a1c` · `domain:cli` seat 1 · `zhuangjianguo` · session `session_01SYEJqDL3NXxrJRBVqaMu9T` · 2026-10-18T22:59Z

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

os serve --port 0 prints "WARN Port 0 is in use, using port N instead" at the default log level — a kernel-assigned port is a legal request, not a busy port (plugin-hono-server's kernel:listening compare) #24387

Description

What happens

Where

Expected

Dedupe

Activity

objectstack-fleet commented on Oct 18, 2026

Triage: p3 · domain:cli · pm:queue · triage seat · session_01AavokzJ5DndAwitDXvKy4U · 2026-10-18T19:57Z

objectstack-fleet commented on Oct 18, 2026

objectstack-fleet commented on Oct 18, 2026

objectstack-fleet commented on Oct 18, 2026

objectstack-fleet commented on Oct 18, 2026

Landed: PR #24402 → 2a1e48a1c · domain:cli seat 1 · zhuangjianguo · session session_01SYEJqDL3NXxrJRBVqaMu9T · 2026-10-18T22:59Z

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

Triage: p3 · `domain:cli` · `pm:queue` · triage seat · `session_01AavokzJ5DndAwitDXvKy4U` · 2026-10-18T19:57Z

Landed: PR #24402 → `2a1e48a1c` · `domain:cli` seat 1 · `zhuangjianguo` · session `session_01SYEJqDL3NXxrJRBVqaMu9T` · 2026-10-18T22:59Z