…wenLM#4466) (QwenLM#4474)

* fix(config): load home .env vars before settings ${VAR} resolution (QwenLM#4466)

${VAR} placeholders in settings.json (e.g. MCP server headers) could not
reference variables defined in ~/.qwen/.env because resolveEnvVarsInObject()
ran before loadEnvironment() loaded the .env file into process.env.

Add preLoadHomeEnvVars() that loads all variables from home-level .env files
(~/.qwen/.env, ~/.env) into process.env in no-override mode before settings
env var resolution. Workspace .env files and settings.env are still handled
by the existing loadEnvironment() call after settings merge.

Fixes QwenLM#4466

* test: add env var resolution from .env file test (QwenLM#4466)

* fix(config): use customEnv fallback for home .env resolution (QwenLM#4466)

Co-Authored-By: Claude Opus 4 (1M context) <[email protected]>
Signed-off-by: kagura-agent <[email protected]>

* fix: align .env precedence and fix test mock

- Use ??= in getHomeEnvFallbackVars() so the more-specific file
  (~/.qwen/.env) wins over ~/.env, matching dotenv first-occurrence-wins
- Fix test 'QWEN_HOME is set': use customSettingsPath matching the
  runtime QWEN_HOME instead of the module-level USER_SETTINGS_PATH
  constant (fixes mock mismatch that cascaded into 7 test failures)
- Remove unused homeEnvPath variable (tsc/eslint warning)
- Add debugLogger.warn in catch block for I/O errors
- Document that the dict intentionally skips
  PROJECT_ENV_HARDCODED_EXCLUSIONS (substitution scope is narrower
  than process.env population)

Signed-off-by: kagura-agent <[email protected]>

* test: add .env fallback, precedence, and error path tests

Address R2 reviewer suggestions:
- Add ~/.env fallback positive path test (no QWEN_HOME)
- Add precedence test: ~/.qwen/.env wins over ~/.env (first-write-wins)
- Add error path test: readFileSync throws, loadSettings still succeeds
- Add code comment explaining intentional path discrepancy between
  getHomeEnvFallbackVars() and getUserLevelEnvPaths()

* docs: clarify customEnv precedence comment per reviewer suggestion

---------

Signed-off-by: kagura-agent <[email protected]>
Co-authored-by: Claude Opus 4 (1M context) <[email protected]>

* fix(core): enforce adjacent tool results

* fix(core): handle orphan cleanup edge cases

* fix(cli): hide completed sticky todos

* fix(cli): remeasure sticky todos on status changes

* feat(cli): add settings corruption recovery dialog

* fix(cli): address review1 comments – copy instead of rename, clean DEBUG logs & lint

* test(cli): add SettingsCorruptedDialog keyboard navigation & callback coverage

* Update packages/cli/src/config/settings.test.ts

Co-authored-by: Shaojin Wen <[email protected]>

* Update packages/cli/src/config/settings.ts

Co-authored-by: Shaojin Wen <[email protected]>

* fix(cli): drop corrupted-path filter, use corruptedPath as sole signal

* fix(test): add missing truncateToItem to mockUIState

* fix(cli): prevent env var stale state and unnecessary normalization writes

* Apply suggestions from code review

Co-authored-by: Shaojin Wen <[email protected]>

* fix(cli): resolve merge conflict and fix formatting errors from PR review update

Co-authored-by: Shaojin Wen <[email protected]>

* fix(cli): prevent validateAuthMethod from consuming corruption env vars

* test(settings): fix scope guard test to actually set env vars

* fix(cli): separate corruption warning from migrationWarnings

* fix(ui): extract CORRUPTED_SUFFIX constant to avoid magic string

* fix(ui): emit error on restore failure in onExit handler

* test(ui): strengthen up/down arrow test to assert selection on target line

* test(settings): strengthen double-corruption and scope guard assertions

* test(ui): assert selection moves in onContinue test before pressing Enter

* fix(ui): include corruption dialog in dialogsVisible to suppress global keypress

* fix(ui): resolve type incompatibility in dialogsVisible and test helpers

* fix(cli): block corruption env var injection and harden error handling

Co-Authored-By: wenshao <[email protected]>

* fix(cli): harden corruption env var guard with helper, comment, and regression test

* fix(cli): add afterSpawn callback to clear corruption env vars after spawn

Co-Authored-By: qwen3.7-max <[email protected]>

* fix(cli): share mockSpawn instance between default and named exports in relaunch test

* Update packages/cli/src/utils/relaunch.ts

Co-authored-by: Shaojin Wen <[email protected]>

---------

Co-authored-by: Shaojin Wen <[email protected]>
Co-authored-by: qwen3.7-max <[email protected]>

)

The Auto-memory / Auto-dream / Auto-skill rows initialized their state
from Config getters, which are frozen at startup and never reflect a
setValue() write. Each /memory reopen re-mounts the dialog and re-reads
that stale snapshot, so a just-flipped toggle appeared to revert. Read the
initial state from the live merged settings instead, matching the existing
write path (bareMode semantics preserved).

Also switch the test's `act` import to `react` — the previously used
@testing-library/react is declared in package.json but not installed, so
the suite could not run — and add a mount/unmount/remount regression test.

…rite files (QwenLM#4431)

* fix(core): preserve uid/gid in atomicWriteFile to avoid breaking shared-write files

atomicWriteFile uses write-to-tmp + rename for crash atomicity. POSIX
rename creates a new inode owned by the calling process's euid/egid, so
the rename silently strips the original uid/gid. On shared-write setups
(e.g. a group-writable file owned by another user in a shared workspace
where the current user has group-write access), every Write/Edit/
NotebookEdit through qwen-code would reset ownership to the running
user and effectively revoke write access for the original collaborators.

The fix:

1. If the target exists and is owned by a different uid/gid than the
   process's effective uid/gid (and we are not root), fall back to
   in-place writeFile. This truncates the existing inode in place,
   preserving uid/gid. The trade-off is loss of crash atomicity for
   this specific case — an acceptable trade for not silently breaking
   shared-write file ownership.

2. If running as root, atomic rename is still used, and ownership is
   restored via chown(uid, gid) after the rename. Root can chown back;
   non-root cannot, hence the in-place fallback for non-root.

3. Windows is unaffected (no POSIX ownership semantics).

Tests:

- New: in-place fallback on uid mismatch — verify content updates, mode
  preserved, and inode unchanged (the inode is the signal that the
  fallback path ran rather than rename).
- New: same scenario triggered via gid mismatch.
- New: positive case — ownership matches → atomic rename → inode changes.

Regression: a v0.16.0 user reported "every write turns a world-writable
file into one other users can no longer write." Bisected to QwenLM#4096 which
introduced atomicWriteFile + write-to-tmp + rename.

* fix(core): route root through in-place fallback + doc/test follow-ups

Review follow-ups on the atomic-write ownership fix:

1. Remove the root-special-case (rename + post-rename chown). chown
   silently fails inside user-namespaced or CAP_CHOWN-stripped Docker
   containers, which re-triggers the original bug for root-in-Docker
   users — exactly the scenario this fix was reported against. Routing
   root through the same in-place fallback as non-root eliminates this
   failure mode and drops an untestable branch (chown-back can't be
   exercised under non-root CI).

2. Document the three properties traded away by the in-place fallback:
   crash atomicity, concurrent-reader isolation, inotify watcher
   semantics (MODIFY vs MOVED_TO).

3. Document that the in-place fallback surfaces EACCES when the file's
   mode forbids the current user from writing — this is correct
   behavior (atomic rename used to silently replace files the user had
   no permission on, which was arguably a privilege issue).

4. Replace the brittle "see step 6 in the function doc" comment with a
   step-number-independent reference.

5. New test covering the EACCES path: chmod 0o444 + mocked geteuid
   triggers the fallback, fallback hits the read-only file, EACCES
   propagates cleanly, original content is preserved.

* fix(core): harden in-place fallback against symlink/unlink/inode races + doc/test follow-ups

Review follow-ups on QwenLM#4431 ownership-preservation fix:

CRITICAL — in-place fallback security hardening (wenshao review):

The path-based `fs.writeFile(targetPath, ...)` fallback introduced
three races that the prior `rename(tmp, target)` form did not have:

1. Non-regular files (FIFO/socket/device): fs.writeFile calls
   open(O_WRONLY|O_CREAT|O_TRUNC). On a FIFO this blocks forever
   waiting for a reader. On a character/block device it writes to
   the actual device. The rename path replaced these with a
   regular file.

2. Symlink-swap TOCTOU: an attacker with parent-dir write can swap
   targetPath for a symlink between our stat and our writeFile.
   fs.writeFile follows symlinks at the destination; POSIX rename
   does not. In the very "shared-write workspace / Docker bind-mount"
   scenarios this PR targets, this lets a directory-writable
   attacker redirect agent writes elsewhere (e.g. /etc/passwd if
   the agent runs as root).

3. Unlink race: if targetPath is unlinked between stat and write,
   O_CREAT silently recreates it owned by the calling user — the
   exact ownership change the fallback was designed to prevent.
   Silent regression to the pre-fix bug under this race.

Fix: extract the fallback into writeInPlaceWithFdGuards():

  - open(target, O_WRONLY | O_TRUNC | O_NOFOLLOW) — no O_CREAT, so
    unlink-race surfaces ENOENT instead of silently recreating; and
    O_NOFOLLOW rejects symlink-swaps with ELOOP.
  - fstat(fd) verifies the bound inode's uid/gid still match
    existingStat — refuses the write if an inode-swap happened
    between stat and open.
  - Write through the fd (locked to the verified inode), chmod
    through the fd, close.

Caller now gates the fallback on existingStat.isFile() — non-regular
targets fall through to the atomic path which has well-defined
"replace special-file with regular-file" semantics.

DOC / TEST follow-ups:

- Add hardlink-propagation as a 4th trade-off in the in-place
  fallback JSDoc (review comment QwenLM#4): rename creates a new inode so
  sibling hardlinks keep old content; in-place truncate+write keeps
  the inode so all hardlinks see new content.

- Update atomicWriteJSON JSDoc to note the write is now
  *conditionally* atomic (review comment QwenLM#5): atomic when uid/gid
  matches the process, in-place when ownership differs. Previously
  the JSDoc still claimed unconditional atomicity.

- Update caller comments at runtimeStatus.ts and
  worktreeSessionService.ts that advertised crash-atomic writes via
  tmp+rename — those guarantees are now conditional (review
  comment QwenLM#6).

- Add mode + tmp-leftover assertions to the gid-mismatch test to
  match the uid-mismatch test (review comment #2 — test
  consistency). Without these, a gid-fallback regression that
  silently dropped permissions or left a tmp file would not be
  caught.

- New test: FIFO + ownership mismatch must take the atomic path,
  not in-place (verifies the existingStat.isFile() guard works;
  hang on in-place would trip vitest timeout).

- New test: writing through a symlink with ownership mismatch
  exercises the resolve-then-stat-then-open flow and verifies the
  symlink itself is preserved.

Tests: 192/192 pass (atomicFileWrite + write-file + edit +
fileSystemService).

* fix(core): defer O_TRUNC and verify dev+ino in writeInPlaceWithFdGuards

PR QwenLM#4431 review follow-up (wenshao critical):

The previous form opened with `O_WRONLY | O_TRUNC | O_NOFOLLOW`, which
truncated the bound file *before* the fd-bound fstat verification ran.
If an attacker swapped the path between the caller's stat and our
open, we would truncate the attacker's substituted inode (destroying
unrelated content) before detecting the swap.

Two fixes:

1. Open without O_TRUNC. Verify dev+ino+uid+gid+isFile match
   expectedStat through fh.stat(). Only then call fh.truncate(0)
   through the validated fd.

2. Expand the verification beyond uid+gid to include dev+ino+isFile.
   uid+gid alone misses a same-owner inode swap (attacker replaces
   the path with a different inode they own). dev+ino is the strong
   identity check; isFile catches a swap to FIFO/socket/device after
   the caller's existingStat.isFile() gate.

JSDoc updated to enumerate the four guards (NOFOLLOW, no CREAT, no
TRUNC at open, dev+ino+uid+gid+isFile via fstat) and explain why
truncation must wait until after verification.

192/192 tests pass.

* fix(core): close FIFO swap race with O_NONBLOCK + cover EOWNERSHIP_CHANGED path

PR QwenLM#4431 review follow-up (deepseek-v4-pro via /review):

CRITICAL — FIFO swap TOCTOU:

The caller's `existingStat.isFile()` gate uses stat data captured
earlier. An attacker with parent-dir write can swap the regular file
for a FIFO between the caller's stat and our open inside
`writeInPlaceWithFdGuards`. The previous `O_WRONLY | O_NOFOLLOW` open
would then block indefinitely waiting for a FIFO reader; O_NOFOLLOW
only catches symlinks.

Fix: add O_NONBLOCK to the open flags. Defense in depth:

- On a reader-less FIFO, `open(O_WRONLY | O_NONBLOCK)` returns ENXIO
  immediately — no hang.
- If the FIFO has a reader (open succeeds), the subsequent fstat
  isFile() check still refuses the write via EOWNERSHIP_CHANGED.
- For regular files, O_NONBLOCK is a no-op.

CRITICAL test gap — EOWNERSHIP_CHANGED branch untested:

The primary TOCTOU defense (fdStat dev/ino/uid/gid/isFile vs
expectedStat) had no coverage. Exported `writeInPlaceWithFdGuards` so
it can be unit-tested directly:

- New test: simulate post-stat inode swap (unlink + recreate at same
  path), call helper with stale stat, assert EOWNERSHIP_CHANGED and
  that the attacker's content survives.
- New test: simulate post-stat regular→FIFO swap, assert open fails
  fast (ENXIO) or fstat catches it — either way no hang, no write.

DOC fix:

JSDoc said "we open read-write without truncating" but the code uses
O_WRONLY. Wording corrected to "write-only".

194/194 tests pass.

* fix(core): fix flaky inode-swap test + apply review follow-ups

PR QwenLM#4431 review follow-up (glm-5.1 via /review) — 7 suggestions adopted,
1 partially adopted, 0 rejected:

CI FIX (Ubuntu test failure on tmpfs inode reuse):

The EOWNERSHIP_CHANGED inode-swap test used unlink+create to simulate
a post-stat swap. On Linux tmpfs the freshly-freed inode number is
often reused by the immediately-following create, so dev+ino remained
identical and the guard didn't trip (intermittent on Ubuntu CI; macOS
APFS happened to allocate different inodes). Switched to rename(decoy,
target) which moves an existing distinct inode into place, guaranteed
to differ from the original.

CODE:

- Wrap fh.writeFile failure after fh.truncate(0) with
  EINPLACE_WRITE_FAILED + cause, so callers see explicitly that the
  file was truncated and the write didn't complete (otherwise they
  see raw ENOSPC/EIO and may wrongly assume the original is intact
  given this lives in atomicFileWrite.ts).
- Skip fh.chmod when euid is neither root nor expectedStat.uid —
  chmod is guaranteed to fail with EPERM in that case (POSIX requires
  owner or root). Avoids a guaranteed-failing syscall on every call.
- Caller catches ENOENT from writeInPlaceWithFdGuards and falls
  through to atomic rename path. If the file was deleted between
  caller's stat and our open there is no ownership to preserve; the
  rename path correctly creates a new file at targetPath.

DOC:

- Replaced "defends against four races" with "hardened against
  post-stat races" (the bullet list has 5 items, the count was wrong).
- Reworded "non-regular targets must not reach this function" to
  describe defense-in-depth — O_NONBLOCK + !fdStat.isFile() reject
  post-stat regular→FIFO/socket/device swaps. The old wording made
  it look like O_NONBLOCK was redundant.
- Documented the dual chmod behavior (root vs non-root with foreign
  uid) inline.

TESTS:

- Added happy-path test for writeInPlaceWithFdGuards (write succeeds,
  inode preserved, mode preserved).
- Added ENOENT regression test (verifies the missing-O_CREAT
  property — if file unlinked between stat and open, no silent
  recreate with caller's uid).
- Renamed the misleading "O_NOFOLLOW guard" test (it actually tests
  resolve-through-symlink, not O_NOFOLLOW) to reflect what it does,
  and added a direct ELOOP test that drives writeInPlaceWithFdGuards
  with a path whose final component is a symlink — that's the real
  O_NOFOLLOW exercise.
- Fixed the FIFO test to pass a stat captured from the FIFO itself
  (not a stale regular-file stat) so only the FIFO-specific defense
  fires, not the inode/dev mismatch from a different file.

NOT ADOPTED:

- Skip-when-non-root chmod optimization adopted (small, useful), but
  the larger "structured chmod error model" deferred — best-effort
  matches the existing tryChmod pattern at file scope.

197/197 tests pass.

* fix(core): wrap truncate err + post-write nlink check + guard close + chmod sync

PR QwenLM#4431 review follow-up (qwen-latest-series-invite-beta-v34 via /review)
— 7 of 10 suggestions adopted, 3 deferred:

CODE:

- **EINPLACE_TRUNCATE_FAILED wrap** (review #3291863048): symmetric to
  the existing EINPLACE_WRITE_FAILED — distinguishes "truncate failed,
  original intact" from "write failed post-truncate, original lost".

- **Post-write nlink === 0 check** (review #3291863059):
  EINODE_UNLINKED_DURING_WRITE detects the fstat-to-close window where
  a concurrent rename-over drops our bound inode's link count to zero
  and our write goes to an anonymous inode close will free. Silent
  data loss path now surfaces.

- **fh.close() guarded in finally** (review #3291863044): close failure
  on NFS/FUSE was masking the original try-body exception (including
  the meaningful EOWNERSHIP_CHANGED, EINPLACE_*, EINODE_*). flush:true
  already fsync'd, so close-after-flush is best-effort.

- **fdStat.uid in canChmod** (review #3291863055 part 1): use the
  fd-bound verified value instead of expectedStat.uid. Defense in depth
  — a future weakening of the fstat guard won't silently widen chmod
  privilege.

- **fh.sync() after chmod** (review #3291863053): chmod is metadata,
  not covered by writeFile({ flush: true }). A crash before lazy
  metadata flush would lose the mode restoration (matters for
  setuid/setgid). One extra syscall, best-effort.

- **@remarks freshness contract** (review #3291863051 partial): JSDoc
  now spells out that expectedStat MUST be a fresh stat captured
  immediately before the call. Stale stats nullify every guard.

- **Concurrent-writer limitation noted** (review #3291863061 partial):
  added a "Known limitation — no advisory locking" paragraph to JSDoc
  rather than adopting flock (Linux-specific, NFS issues, scope
  expansion). Callers needing multi-process coordination should layer
  their own lockfile.

- **@throws documentation** (review #3291863051 partial): four
  documented error codes (EOWNERSHIP_CHANGED, EINODE_UNLINKED_DURING_WRITE,
  EINPLACE_TRUNCATE_FAILED, EINPLACE_WRITE_FAILED).

TESTS:

- **EINPLACE_WRITE_FAILED via FileHandle.prototype.writeFile monkey-patch**
  (review #3291863040): triggers the data-loss path, asserts the wrapped
  code + message + cause, and verifies the file is empty (truncate ran).

- **canChmod=false actually skips chmod** (review #3291863055 part 2):
  prior uid-mismatch test had desiredMode === current mode, couldn't
  distinguish "skipped" from "no-op". New test uses desiredMode=0o755
  on a 0o644 file under canChmod=false → asserts mode stays 0o644.

NOT ADOPTED:

- ENOENT/ELOOP/ENXIO catch extension (review #3291863043): keeping the
  strict refusal for swap-to-special-file. Silent fallthrough-to-replace
  was pre-PR atomic-rename behavior, but in shared-write workspaces
  (this PR's target users) a special-file appearing at the target path
  is a signal worth surfacing, not papering over.

- Diagnostic logging (review #3291863049): the function has no logger
  dependency today; adding one is an architecture decision outside
  this PR's scope. The path taken is implied by the side effects
  (inode preserved vs new) but agreed: out-of-band telemetry would
  help ops. Defer to follow-up.

- flock advisory locking (review #3291863061 main): scope expansion;
  Linux-specific semantics, NFS edge cases. Documented as known
  limitation instead.

- Integration test for ENOENT fallthrough at atomicWriteFile level
  (review #3291863043 part 1): ESM module bindings prevent monkey-
  patching writeInPlaceWithFdGuards from outside. The unit test for
  the helper's ENOENT path covers the throwing behavior; the catch is
  3 lines and review-visible. Defer until a refactor opens an
  injection seam.

- Error code string constants export (review #3291863051 part 3): two
  codes don't merit a constant module. Magic strings are fine at this
  size.

199/199 tests pass.

* docs(core): sync writeRuntimeStatus JSDoc with conditional-atomic contract

PR QwenLM#4431 review follow-up: function-level JSDoc still claimed
unconditional "Atomically write" and "never sees a partially written
file", inconsistent with the module-level docblock updated in earlier
commits. Updated to describe the conditional-atomic behavior (atomic
when uid/gid matches, in-place fallback when ownership differs) and
explicitly note the concurrent-reader visibility trade-off in the
fallback path. Links to atomicWriteJSON for the full contract.

Doc-only change. 199/199 tests pass.

* fix(core): add explicit fh.sync() — FileHandle.writeFile ignores flush option

PR QwenLM#4431 review follow-up (qwen3.7-max via /review):

CRITICAL — FileHandle.writeFile silently ignores flush:

Node.js FileHandle.writeFile takes an early-return path that bypasses
the flush option entirely (the option is only honored on the
path-based fs.writeFile form). Our previous code passed
{ flush: true } to fh.writeFile and relied on the implicit fsync.
The only explicit fh.sync() was nested in the chmod block guarded by
canChmod — which is FALSE precisely when a non-root group member
writes to a group-writable file they don't own (the exact shared-write
scenario this PR targets). Net effect: in that branch, zero fsync.
Data sits in the kernel page cache; a crash before lazy flush leaves
the file empty (truncate succeeded) or partially written.

Fix:
- Drop flush from the fhWriteOptions object (silently ignored anyway).
- Add an explicit `fh.sync()` after writeFile succeeds, gated on
  options.flush. Runs BEFORE the chmod block so the canChmod=false
  branch also fsyncs.
- The chmod-block fh.sync() becomes metadata-only (covers the mode
  change), as the data is already on disk.

Updated comments to reflect the actual semantics rather than the
incorrect "writeFile({ flush: true }) fsyncs" assumption.

TESTS (partial adoption of review #3293252349):

- EINPLACE_TRUNCATE_FAILED: sibling test to EINPLACE_WRITE_FAILED.
  Monkey-patches FileHandle.prototype.truncate to throw EIO; asserts
  err.code + cause + "original content is intact" message, and
  verifies the file's original bytes are unchanged (truncate didn't
  run).
- Buffer in in-place fallback: locks in binary fidelity (byte-exact
  comparison) so a future encoding-passthrough regression for Buffer
  data would be caught.

NOT ADOPTED in this commit:

- EINODE_UNLINKED_DURING_WRITE test: requires post-write fh.stat()
  mocking with call-count discrimination (first call: real stat for
  verification; second call: nlink=0). The monkey-patch pattern works
  but is fragile; deferred to a follow-up that may also refactor the
  helper to accept an injectable stat fn for cleaner testability.

201/201 tests pass.

* fix: correct stale flush comment + add fh.sync() regression test

- Fix misleading close() comment that said "flush:true already
  fsync'd" — the explicit fh.sync() does the actual fsync, not the
  flush option (which is silently ignored on FileHandle.writeFile).
- Add regression test verifying fh.sync() is called when flush:true
  and skipped when flush is absent, preventing silent removal of the
  core durability fix.

Addresses wenshao review threads from 2026-05-23.

* test: add EINODE_UNLINKED_DURING_WRITE regression test

Monkey-patches FileHandle.stat to return nlink:0 on the post-write
check, verifying the nlink guard throws with the correct error code.
Addresses wenshao review from 2026-05-28.

* simplify: replace writeInPlaceWithFdGuards with plain fs.writeFile

Address yiliang114's review (CHANGES_REQUESTED):

1. [Critical] Remove ~120 lines of fd-level TOCTOU hardening
   (writeInPlaceWithFdGuards) — over-engineering for a local CLI.
   The in-place fallback now uses plain fs.writeFile + tryChmod,
   matching the EXDEV fallback pattern.

2. [Suggestion] Fix macOS GID false-positive: only compare uid in
   ownershipWouldChange(). macOS inherits parent dir GID for new
   files, so egid !== file.gid was a false positive that needlessly
   dropped crash atomicity.

3. [Suggestion] Trim 60+ lines of JSDoc to project style (AGENTS.md:
   "default to none, add only when WHY is non-obvious").

Net: -748 lines. 24 tests pass.

* fix: restore Stats type import (TS2304 build failure)

* docs: narrow scope from uid/gid to uid-only preservation

The gid check is intentionally skipped because macOS inherits the
parent directory's GID for new files, making egid !== file.gid a
false positive. Update comments and PR description to match the
actual implementation scope.

* test: add inode assertion to symlink ownership-mismatch test

Proves the in-place fallback actually ran instead of atomic rename.

* feat(cli): improve hooks matcher display

* test(cli): cover hooks navigation levels

Detach closeSession/killSession from the session entry's owning channel instead of the current attach target, so the correct channel is decremented and killed during channel overlap (old channel dying while a fresh channel is current). Extracts findChannelInfoForEntry/detachSessionIdFromEntryChannel helpers with unit + integration coverage. Fixes QwenLM#4325.

… variants to prevent OOM on resume (QwenLM#4644)

* fix(core,cli): replace full-history structuredClone with shallow/tail variants to prevent OOM on resume

Several UI and service call sites clone the entire chat history via
structuredClone(getHistory()) every turn. On a resumed session with
thousands of entries, each clone allocates 150-200 MB transiently.
When multiple async side-requests overlap (suggestion generation,
auto-title, checkpointing), multiple clones coexist on the heap,
pushing V8 past its limit within 10 turns (2 GB heap cap).

Changes:
- AppContainer.tsx: use getHistoryTail(40, true) instead of
  getHistory(true) + slice(-40)
- btwCommand.ts: same pattern, use getHistoryTail(40, true)
- sessionTitle.ts: use getHistoryShallow() (read-only filtering)
- sessionRecap.ts: use getHistoryShallow() (read-only filtering)
- useGeminiStream.ts: use getHistoryShallow() for checkpoint
  serialization (only needs to survive JSON.stringify)

Closes QwenLM#4624

* fix(test): update mocks for getHistoryShallow/getHistoryTail in sessionTitle and btwCommand tests

* fix(cli): migrate remaining getHistory() clone sites to shallow/tail variants

- AppContainer.tsx rewind path: getHistory() → getHistoryShallow()
  (only used read-only by computeApiTruncationIndex)
- Session.ts ACP rewind: getHistory() → getHistoryShallow()
  (only walks entries to compute truncation index)
- Session.ts stop-hook: getHistory() + filter(.model).pop() →
  getLastModelMessageText() (O(1) backward scan, no clone)

* fix(core): use client-level getHistoryShallow with fallback

sessionTitle.ts and sessionRecap.ts were calling
chat.getHistoryShallow() directly, bypassing the client-level
wrapper that provides a getHistory() fallback when the chat
implementation doesn't support shallow reads. Use
geminiClient.getHistoryShallow() instead.

Update test mocks to match the new call site.

* fix(test): add getHistoryShallow and getLastModelMessageText to Session test mocks

Session.ts now calls chat.getHistoryShallow() in rewindToTurn and
chat.getLastModelMessageText() in the Stop hook. Update all mockChat
instances in Session.test.ts to provide these methods.

… statusline (QwenLM#4670)

* feat(cli): add respectUserColors option to preserve ANSI colors in
     statusline command output

* test(cli): add respectUserColors tests for useStatusLine and Footer

* feat(cli): add hideContextIndicator option to hide built-in context usage in footer

* docs: update statusline configuration docs with respectUserColors and hideContextIndicator

Fixes QwenLM#4326

…e handling (QwenLM#3557)

* Harden insight facet normalization and empty qualitative handling

* feat: enhance AtAGlance component to accept target sections for dynamic rendering

* feat(core): add simplify bundled skill

Co-authored-by: Qwen-Coder <[email protected]>

* test(cli): stabilize SettingsDialog restart prompt test

Co-authored-by: Qwen-Coder <[email protected]>

* fix(skills): use agent tool instead of task in simplify skill

The simplify skill referenced the 'task' tool for launching review passes,
but Qwen Code exposes 'agent' as the callable subagent tool ('task' is only
a legacy permission alias). Using 'task' would cause /simplify to stall when
trying to launch parallel review passes.

Co-authored-by: Qwen-Coder <[email protected]>

* docs: document simplify bundled skill

Co-authored-by: Qwen-Coder <[email protected]>

* Update packages/core/src/skills/skill-manager.test.ts

Co-authored-by: Shaojin Wen <[email protected]>

* fix(core): repair simplify skill tests

Co-authored-by: Qwen-Coder <[email protected]>

* Update packages/core/src/skills/bundled/simplify/SKILL.md

Co-authored-by: Shaojin Wen <[email protected]>

* fix(skills): address simplify review feedback (read-only passes, gitignore scope, safer dead-code removal)

- drop inert `argument-hint` frontmatter (argumentHint is never parsed or
  rendered anywhere; no other bundled skill uses it)
- mark Step 2 review passes read-only so edits stay isolated to Step 4
- narrow the no-diff fallback to `git ls-files --modified --others
  --exclude-standard` so ignored build output is excluded
- require a repo-wide caller check before removing code
- make the commands.md row state it edits code directly
- assert non-conflicting bundled skills survive cross-level dedup

Co-authored-by: Qwen-Coder <[email protected]>

---------

Co-authored-by: Qwen-Coder <[email protected]>
Co-authored-by: Shaojin Wen <[email protected]>
Co-authored-by: wenshao <[email protected]>

* chore(skills): add codex reproduce workflows

* feat(agent-reproduce): implement agent reproduction workflow and supporting scripts

* feat(skills): capture reference agent state diffs

)

* chore(deps): re-upgrade ink 6 → 7.0.3 (upstream Static remount fix landed)

PR QwenLM#3860 first upgraded ink 6 → 7.0.2. PR QwenLM#4083 reverted because of a
TUI regression: `<Static>` did not re-emit items when its `key` prop
was bumped, so `/clear` / Ctrl+O / refreshStatic left the history area
blank under ink 7.0.2.

ink 7.0.3 (released after QwenLM#4083) contains the exact fixes:

  - be9f44cda Fix: <Static> remount via key change drops new items (QwenLM#948)
  - 669c4386c Fix: Drop stale <Static> output from fullStaticOutput on identity change (QwenLM#950)
  - 7c2267c01 Fix `useBoxMetrics` not accepting ref objects with an initial null value (QwenLM#945)

Changes:
  - `ink` ^6.2.3 → ^7.0.3 (root hoist + cli direct)
  - `react` ^19.1.0 → ^19.2.4 (cli direct; ink 7.0.3 peerDeps requires >=19.2.0)
  - `react`/`react-dom` overrides ^19.2.4 added so the transitive graph
    stays deduped to a single instance (avoids `Invalid hook call` from
    multiple React copies, the classic ink-upgrade hazard)
  - `wrap-ansi` already on ^10.0.0 from QwenLM#4083's partial-revert (no change)

Verified:
  - `npm ls ink` → single `[email protected]` across all peer deps
  - `npm ls react` → single `[email protected]`
  - `npm run typecheck --workspace=@qwen-code/qwen-code` clean
  - `npm run typecheck --workspace=@qwen-code/qwen-code-core` clean
  - Composer.test.tsx 20/20, MainContent.test.tsx 6/6, TableRenderer.test.tsx
    59/59 + 1 skipped — all key UI components green on the new ink

The Static-remount regression is upstream-fixed in 7.0.3, so the
runtime path is restored without needing QwenLM#3941's overflowY-self-managed
viewport. QwenLM#3941 (virtual viewport) remains an opt-in performance
feature on top.

* fix(deps,cli): add @types/react overrides + move refreshStatic out of setCurrentModel updater

Two follow-ups from the multi-round audit of the ink 7.0.3 re-upgrade:

1. @types/react / @types/react-dom now pinned to ^19.2.0 in root
   overrides. packages/web-templates still declares @types/react ^18.2.0
   in its devDeps. Today the CLI build is unaffected (web-templates's
   18.x types are nested in its own node_modules and the React-using
   src/insight and src/export-html files are excluded from its tsconfig
   build), but a future reincludes-or-hoist accident would land
   conflicting global JSX namespaces in the CLI compile graph. Match
   the dep dedup we already enforce for `react` and `react-dom` so the
   type graph stays as deduped as the runtime graph.

2. AppContainer's onModelChange handler was calling refreshStatic() as
   a side-effect inside the setCurrentModel updater. React.StrictMode
   double-invokes state updaters in dev, so model swaps fired two
   clearTerminal writes + two <Static> key bumps. The double work was
   masked under ink 6 (key changes were no-ops on <Static>), but ink
   7.0.3 honors key changes — the doubled work is now potentially
   visible as a faster flash-flash on every model switch.

   Refactor: setCurrentModel becomes a pure setter; refreshStatic
   moves into a useEffect keyed on currentModel with a ref-comparison
   guard so the first render doesn't fire. Single clearTerminal write
   per real model change, even under StrictMode.

Verified: npm ls ink → single 7.0.3, npm ls react → single 19.2.4,
npm ls @types/react → 19.2.10 hoisted (npm flags web-templates's 18.x
constraint as overridden, which is the intended behavior). Typecheck
clean across cli + core workspaces.

* docs(design): virtual viewport on ink 7 — analysis + PR sequence

Captures the architectural analysis of how to thoroughly close the
flicker / refresh-storm class of issues (QwenLM#2950, QwenLM#3118, QwenLM#3007, QwenLM#3838 UI
side, QwenLM#3899 follow-on) using a virtualized history viewport.

- Surveys claude-code (forked ink) and gemini-cli (@jrichman/ink +
  ScrollableList + VirtualizedList) reference implementations.
- Confirms ink 7 already exposes the primitives needed
  (`useBoxMetrics`, `measureElement`, `useWindowSize`,
  `useAnimation`) — no fork swap required.
- Picks porting gemini-cli's virtualized list components to ink 7 with
  `ResizeObserver` -> `useBoxMetrics` and a custom `StaticRender`.
- Splits the work into V.0..V.4 PRs with scope, dependencies, risk.
- Lists open questions + 11-item approval checklist that must clear
  before V.0 implementation begins.

This is a docs-only PR per the project's design-first workflow. No
runtime code changes.

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

* feat(cli): virtual viewport for long conversations on ink 7

Port gemini-cli's VirtualizedList + ScrollableList to stock ink 7,
adapting for ink 7's available primitives:

- `overflowY="hidden"` + `marginTop={-scrollTop}` instead of ink-fork's
  `overflowY="scroll"` (ink 7 has proper clip/unclip in render-node-to-output)
- `useBoxMetrics` inside each VirtualizedListItem (Option A) instead of a
  single ResizeObserver WeakMap; reports height changes via onHeightChange
  callback so the parent can update its heights record
- Custom `StaticRender` as `React.memo` with a reference-equality comparator,
  keyed on `itemKey-static-{width}` to freeze completed conversation items
- Character scrollbar column (`│` track / `█` thumb) since ink 7 has no
  native scrollbar prop
- No ScrollProvider / mouse drag (deferred to a follow-up PR)

Wire into MainContent.tsx behind `ui.useTerminalBuffer` setting (Settings
dialog → UI → Virtualized History; default false — opt-in).

Key bindings: Shift+↑/↓ (line), PgUp/PgDn (page), Ctrl+Home/End (top/bottom).

Re-render optimisations:
- renderItem wrapped in useCallback so renderedItems useMemo only recomputes
  when actual deps change (not on every streaming tick)
- Completed history items passed by original object reference so
  VirtualHistoryItem = memo(HistoryItemDisplay) can bail out on stable props
- estimatedItemHeight / keyExtractor / isStaticItem defined as module-level
  constants with no closure deps

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

* test(cli): add test coverage for virtual viewport scroll bindings and settings

- keyMatchers.test.ts: 6 new test cases for SCROLL_UP/DOWN, PAGE_UP/DOWN,
  SCROLL_HOME/END commands (41 tests total)
- settingsSchema.test.ts: assert ui.useTerminalBuffer is boolean, default false,
  showInDialog true, requiresRestart false

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

* feat(cli): use ink 7 native overflow for VP pending items

In VP mode, pending items are rendered inside VirtualizedList's
overflowY="hidden" container, which uses ink 7's native clipping
as the viewport guard. Remove the availableTerminalHeight JS-
truncation bound from pending items in renderVirtualItem:

- JS truncation at terminal height would silently cut off content
  the user could scroll to read within the virtual viewport.
- ink 7 overflowY="hidden" on the VirtualizedList container is the
  correct clip guard — no JS line-counting workaround needed.
- Remove uiState.constrainHeight from renderVirtualItem deps (no
  longer referenced in the VP rendering path).

The legacy <Static> path is unchanged.

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

* perf(cli): binary-search offsets in virtualized list hot path

Replace linear findLastIndex / findIndex scans on the offsets array with
upperBound. Offsets are monotonic by construction, so the lookups inside
the render body and getAnchorForScrollTop drop from O(n) to O(log n).
Material for thousand-turn sessions where the lookup runs on every frame.

* fix(cli): wire ShowMoreLines + skip clearTerminal in VP mode

Two audit-found bugs in the VP path:

1. `<ShowMoreLines>` was outside the `<OverflowProvider>` that wraps
   `<ScrollableList>` in VP mode. `useOverflowState()` returns
   `undefined` outside the provider, so the component returned `null`
   and the "press ctrl-s to show more lines" affordance silently
   disappeared. Move `<ShowMoreLines>` inside the provider so the hook
   sees the live overflow state, matching the legacy path.

2. `refreshStatic()` and `repaintStaticViewport()` wrote
   `clearTerminal` / `cursorTo+eraseDown` to the host terminal
   unconditionally. In VP mode the React tree owns the visible region
   via ink 7's native `overflowY="hidden"` clipping — the physical
   write is a wasted flash on Ctrl+O / Alt+M / model change / resize.
   Guard both writes on `useTerminalBuffer === false`. The
   `historyRemountKey` bump still fires so the legacy `<Static>`
   fallback would still remount if someone toggled the setting mid-
   session.

Extends the targeted-repaint pattern introduced in QwenLM#3967 to all
refreshStatic call sites, gated by the VP setting instead of by event
type.

* fix(cli): VP renderItem stability + source-copy offsets + heights GC

Three audit-found regressions tightened, in order of severity:

1. **Source-copy index offsets missing in VP** — legacy `<Static>` path
   threads per-item `sourceCopyIndexOffsets` so `/copy mermaid N` /
   `/copy latex N` hints stay stable across continuation messages. VP
   `renderVirtualItem` was not passing this prop, so the copy hints
   shown under each diagram drifted on every `gemini_content` chunk
   (the clipboard mechanism itself still worked from raw history; only
   the displayed number was wrong). Add two lookup tables —
   identity-keyed for static items, index-keyed for pending — without
   changing the VirtualizedList data signature, and thread offsets in
   both render branches.

2. **`renderVirtualItem` callback invalidated on every streaming tick**
   — its deps included `activePtyId` / `embeddedShellFocused` /
   `isEditorDialogOpen`, all of which flip mid-stream when a shell
   tool runs or a dialog opens. Each flip rebuilt the callback,
   invalidated `VirtualizedList.renderedItems`'s useMemo, and forced
   every static item to re-render through `<StaticRender>` — defeating
   the very memoization the design relies on. Move the three pending-
   only fields into a ref read inside the callback. Static-item closure
   now depends only on inputs that legitimately affect static output
   (terminalWidth, slashCommands, getCompactLabel, …). Pending items
   still re-render correctly because their item identity changes per
   tick, so the callback is called fresh each time and reads the
   latest ref.

3. **`pending` items now honour `constrainHeight`** in VP, matching the
   legacy path. Previously VP unconditionally passed `undefined` for
   `availableTerminalHeight` on pending, relying on the viewport
   `overflowY="hidden"` clip to limit visible size — but that hid the
   `<ShowMoreLines>` affordance from the user. Now that ShowMoreLines
   is correctly wired (previous commit), restore parity.

4. **Heights map memory leak** in `VirtualizedList` — `setHeights` only
   grew. Each `/clear` left orphan `h-N` keys; each pending → completed
   transition left orphan `p-N` keys. Add a `useLayoutEffect` that
   prunes entries whose keys are not in the current `data`. Runs in
   layout phase so the prune commits in the same paint as the data
   change — no stale-offsets frame.

* test+fix(cli): VP path coverage + stabilize absorbedCallIds empty Set

Completion-pass artifacts driven by the multi-agent audit:

- Settings description rewritten to enumerate the symptoms VP fixes so
  users with active flicker reports can find the toggle without reading
  the design doc.
- `absorbedCallIds` returns a module-level constant Set when compact mode
  is off, instead of a fresh `new Set()` per render. Fixes a hidden
  cascade: `activePtyId` flip mid-stream → useMemo runs → returns a new
  empty Set → `isSummaryAbsorbed` rebuilds → `renderVirtualItem`
  rebuilds → `VirtualizedList.renderedItems` recomputes → every static
  item re-renders. With the constant, the cascade dies at the source.
  Helps both VP and legacy paths.
- VP-path unit tests for MainContent (4 cases): ScrollableList mounts
  and Static does not when `useTerminalBuffer: true`; ShowMoreLines is
  reachable in VP mode (regression of the OverflowProvider mis-wrap);
  source-copy index offsets thread into renderItem for static items;
  renderItem callback identity is stable across `activePtyId` flips
  (proves the ref-based read keeps StaticRender memo effective).

* fix(cli): stabilize absorbedCallIds in compact mode + gate heights prune + tighten ShowMoreLines test

Round-2 audit follow-ups. Three real findings addressed; one flagged
false positive documented separately.

1. **absorbedCallIds Set identity now content-stable when compact mode is
   on.** The earlier EMPTY constant only short-circuited the compactMode=
   false path; when compact mode is enabled (some users default-on it),
   activePtyId / embeddedShellFocused flips during streaming still
   produced fresh Sets per render even when membership was unchanged,
   restarting the same cascade the pendingStateRef fix was meant to
   avoid. Compare-and-reuse via a ref: if the new Set has identical
   membership to the previous one, return the previous reference.

2. **`heights` map prune in `VirtualizedList` is gated.** Previously
   every streaming tick rebuilt an N-key Set and walked all heights,
   even on the steady-state path where nothing changes. Now only fires
   when the heights record has clearly outpaced live data
   (`size > max(8, 2 × data.length)`) — covers `/clear` and accumulated
   pending → completed transitions, skips the 30-Hz hot path entirely.

3. **VP ShowMoreLines test now actually verifies overflow connectivity.**
   Previous mock unconditionally rendered "SHOW_MORE", so the test only
   proved the JSX mounted — it would still pass if a future refactor
   moved `<OverflowProvider>` out of the VP tree again. The mock now
   reads `useOverflowState()` and emits "OVERFLOW_DISCONNECTED" when the
   context is missing. The VP test asserts both presence of "SHOW_MORE"
   and absence of the disconnected marker, so the regression is now
   caught.

Not addressed:
- Audit P0-1 claim that `renderMode` (Alt+M) / model-change updates
  don't reach VP static items: false positive. `renderMode` is a React
  Context (`RenderModeContext`), and Context propagation traverses the
  tree past `memo` boundaries — MarkdownDisplay's `useRenderMode()`
  consumer re-renders on context change regardless of whether
  `StaticRender` bails out. Verified by reading
  `packages/cli/src/ui/contexts/RenderModeContext.tsx` and
  `MarkdownDisplay.tsx:172`. No code change.
- Audit P1-2 pendingStateRef write-during-render race: speculative,
  relies on a multi-pass render path React 18+ does not currently use.
  Documented assumption in the existing inline comment.

* fix(cli): isolate renderItem errors + defensive height coerce + compact-mode mergedHistory stability

Round-3 audit follow-ups. Three real findings; the rest verified clean.

1. **`renderItem` errors no longer crash the CLI.** Previously a throw
   inside a per-item render propagated through `VirtualizedList`'s
   useMemo into React's commit phase, tearing down the whole Ink tree —
   one bad history record could nuke the session. Wrap each call in a
   try/catch and substitute a small red `[render error] …` text box on
   failure. The row stays in the viewport so the user can scroll past
   it.

2. **Defensive height coerce in offset accumulation.** A buggy
   `estimatedItemHeight` returning NaN / negative / Infinity would
   poison every downstream offset and break the `upperBound` /
   `findLastLE` binary search (which assumes monotonic offsets). Clamp
   to `Number.isFinite(raw) && raw > 0 ? raw : 0`. No-op for the
   in-tree estimators that return 3; insurance against future
   consumers.

3. **`mergedHistory` is content-stable when compact mode is on.** The
   Round-2 absorbedCallIds stability fix didn't reach this path:
   `mergeCompactToolGroups` always allocates a fresh array, and
   `mergedHistory`'s useMemo lists `activePtyId` / `embeddedShellFocused`
   as deps, so every streaming tick mid-shell-tool produced a new array
   even when items aligned. Cascade went `mergedHistory` → offsets map
   → `renderVirtualItem` → every static item re-rendered. Pair-wise
   compare new vs previous and return the previous reference when items
   align. Restores StaticRender memo effectiveness for compact-mode
   users.

Not addressed (audit findings deemed not worth fixing in this PR):
- `scrollToItem` silently no-ops when item is not in data — no current
  caller checks the return value, low impact.
- `allVirtualItems` array spread is O(n) per streaming tick — real but
  not a crash; revisit in a perf-focused follow-up.
- `itemRefs.current` is dead surface (never read) — cosmetic.
- StrictMode-only-in-DEBUG double-invoke paths verified safe.

* test+chore(cli): VP review round 4 — VirtualizedList/useBatchedScroll coverage + cleanups

Addresses wenshao's CHANGES_REQUESTED review on PR QwenLM#3941.

- Add focused unit tests for `VirtualizedList` (9 cases) covering empty
  data, `renderStatic` full-render, `initialScrollIndex` with
  `SCROLL_TO_ITEM_END`, `targetScrollIndex` anchoring, imperative
  `scrollToEnd` / `scrollToIndex`, per-item `renderItem` error isolation,
  NaN/negative estimator coercion, and out-of-range `initialScrollIndex`
  clamping.
- Add `useBatchedScroll` unit tests (4 cases) covering initial reads,
  pending-value reads in the same tick, post-commit pending reset, and
  callback identity stability across rerenders.
- Remove dead `itemRefs` / `onSetRef` plumbing (declared, written, never
  read; `useCallback` with empty deps was also a stale-closure trap).
- Remove unused `isStatic?: boolean` from `VirtualizedListProps`
  (only `isStaticItem` is actually consumed).
- Tighten the render-phase setState block: each setter is now guarded
  by an equality check so React bails out of redundant updates, and a
  comment documents that this is the React-endorsed "adjusting state
  while rendering" pattern (the synchronous update avoids a one-frame
  flash at the previous position when `targetScrollIndex` changes).

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

* chore(cli): remove dead `dataRef` from VirtualizedList (round-4 followup)

Declared and written in a `useLayoutEffect` on every `data` change but
never read anywhere in the component. Flagged in wenshao's round-4 review
of PR QwenLM#3941.

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

* fix(cli): collapse model-change effect back into one batched handler

wenshao's PR QwenLM#4119 review correctly flagged that splitting the
onModelChange flow into two effects (b25831b) reintroduced the
issue QwenLM#3899 freeze regression on every model switch:

  1. setCurrentModel(model) commits first, with the OLD
     historyRemountKey.
  2. <Static key={`${historyRemountKey}-${currentModel}`}> sees its
     key change (because currentModel did) and remounts immediately.
  3. MainContent's render-phase progressive-replay reset only fires
     when historyRemountKey changes, so replayCount is still the
     full mergedHistory.length from any prior catch-up.
  4. The remounted Static dumps the entire history in one synchronous
     layout pass — exactly the freeze progressive replay was added
     to avoid (QwenLM#3899). The second effect's refreshStatic() bump
     arrives a render too late.

Fix: do not split. Both side effects (refreshStatic, which writes
clearTerminal + bumps historyRemountKey, and setCurrentModel) live
in the event handler again, with a ref guard for same-model
notifications. The React.StrictMode concern that motivated b25831b
is addressed by keeping the side effect OUT of the setState updater
(it now runs once per event-handler invocation, not once per
double-invoked updater call). Both setState calls land in the same
React batch, so historyRemountKey and currentModel update together —
MainContent's render-phase reset sees the new key, replayCount drops
to the first chunk, and Static remounts with chunked replay intact.

Tests:
- AppContainer.test.tsx: 4 new tests covering the synchronous
  refreshStatic side-effect contract, same-model no-op, ref-guarded
  StrictMode double-invoke, and unsubscribe-on-unmount.
- MainContent.test.tsx: new regression guard — when currentModel
  changes but historyRemountKey is held constant, progressive replay
  must NOT reset (pins the MainContent invariant the two-effect
  refactor accidentally relied on).

Verified: vitest packages/cli AppContainer + MainContent green (82/82).
Typecheck clean.

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

* fix+docs(cli): VP review round 5 — typecheck, doc drift, scroll keys

PR QwenLM#4146 review feedback (wenshao + Claude Opus 4.7 audit) addressed:

Code:
- MainContent.test: activePtyId typed as number (was 'pty-xyz' string,
  broke tsc with TS2322 — the test only relies on reference change so
  any number works).
- VirtualizedList: sanitize renderItem error path. Display becomes the
  generic `[render error]` marker; full err goes to debugLogger.debug
  so file paths / partial tool state don't leak to scrollback.
- MainContent: move pendingSourceCopyOffsetsByIndex into a ref so it
  no longer rebuilds renderVirtualItem identity every streaming tick.
  Without this, VirtualizedList.renderedItems useMemo invalidated
  per-tick → JSX rebuilt for every visible item → memo(HistoryItem
  Display) was still bailing but allocations were O(visible) per tick.
- AppContainer: drop the misleading "state-driven scroll reset" claim
  in the VP refreshStatic comment. VP is intentionally near-no-op:
  the React tree owns the visible region, mergedHistory mutation is
  what refreshes the screen, and the remount-key bump is preserved
  only to keep the legacy Static branch in sync if the user toggles
  the flag off mid-session.
- StaticRender: rewrite JSDoc to match reality. The custom React.memo
  is NOT output caching like @jrichman/ink's StaticRender export;
  the comparator rarely matches (parent allocates fresh JSX); the
  real skip happens at memo(HistoryItemDisplay) one level deeper.

Docs:
- docs/design/virtual-viewport: sync file map (drop non-existent
  ScrollProvider.tsx / useAnimatedScrollbar.ts), PR sequence (one PR
  QwenLM#4146, V.3-V.5 deferred), open-question + checklist resolution for
  QwenLM#3905 (superseded) and base branch rename.
- docs/users/reference/keyboard-shortcuts: document the 6 VP scroll
  keys (Shift+↑/↓, PgUp/PgDn, Ctrl+Home/End) under a "History
  scrollback (when ui.useTerminalBuffer is on)" section. Previously
  the only discovery path was the Settings dialog description.

Verified: tsc --noEmit -p packages/cli ✓, vitest 160/160 ✓ across
AppContainer / MainContent / VirtualizedList / useBatchedScroll /
keyMatchers / settingsSchema, eslint clean on touched files.

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

* feat(cli): SGR mouse wheel scroll in VP mode

Recovers the most-felt UX regression vs legacy `<Static>` mode: when
`ui.useTerminalBuffer` is on, legacy users lose mouse wheel as a way
to scroll history (the host terminal stopped seeing the conversation
in its scrollback buffer). This PR enables button-event tracking
(`?1002h`) + SGR coordinates (`?1006h`) while the ScrollableList has
focus, parses wheel events off stdin, and routes them to scrollBy.

Scope kept tight on purpose:
- Wheel only. Hit-testing for scrollbar drag / click-to-position
  needs screen-absolute element coords; stock ink 7's useBoxMetrics
  returns yoga's parent-relative layout. Deferred to V.4 with two
  exit paths (upstream getBoundingBox to ink 7, or local yoga walker).
- Mouse mode is enabled only while ScrollableList is mounted; non-VP
  users never see their terminal flipped into button-event tracking.
- Side effect: native click-and-drag text selection is captured by
  the program. Docs + settings dialog description now spell out the
  Shift / Option (macOS) bypass.

Implementation:
- `ui/utils/mouse.ts` — SGR + X11 parser, ported and trimmed from
  gemini-cli (Google LLC, Apache-2.0). Single-consumer.
- `ui/hooks/useMouseEvents.ts` — enable/parse/disable lifecycle
  hook. Listens on stdin via `useStdin().stdin`, runs handler
  through a ref so callers don't have to memoize.
- `ui/components/shared/ScrollableList.tsx` — subscribe to mouse
  events, route wheel → `scrollBy(±3)`. Also drops a dead outer
  `<Box flexGrow={1}>` wrapper that held an unread containerRef
  and collapsed to zero height in ink-testing-library (the test
  renderer has no flex parent, so flexGrow=1 → 0 height → no items
  ever rendered, which is how this dead code was exposed).

Tests:
- `ui/utils/mouse.test.ts` — 14 cases: SGR parsing (wheel, presses,
  modifiers, move), X11 parsing, fallback chain, incomplete-sequence
  guard (including the >50-byte garbage cap).
- `ui/components/shared/ScrollableList.test.tsx` — 3 cases: wheel
  events shift the rendered window; hasFocus=false makes the mouse
  pipeline inactive (no throw); non-wheel events leave the window
  unchanged. Renders are wrapped in `<KeypressProvider>` (required
  by useKeypress in production but easy to forget in standalone
  tests).

Docs:
- `docs/users/reference/keyboard-shortcuts.md` — adds "Mouse wheel"
  row + the Shift/Option-to-select note.
- `packages/cli/src/config/settingsSchema.ts` — the in-app dialog
  description now mentions mouse wheel and the text-select bypass.
- `docs/design/virtual-viewport/README.md` — §1 status, §5 file map,
  §7 PR sequence all reflect mouse wheel landing in QwenLM#4146 and the
  V.4–V.7 follow-up split (scrollbar drag / in-app search / alt-
  buffer / host-scrollback dual-write research).

Verified: tsc --noEmit -p packages/cli ✓, vitest 182/182 ✓ across
AppContainer / MainContent / VirtualizedList / ScrollableList /
useBatchedScroll / mouse / keyMatchers / settingsSchema.

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

* feat(cli): auto-hide animation for VP scrollbar thumb

Pairs with the SGR mouse-wheel work from the previous commit:
when the user actually scrolls, the thumb pops bright; after a
1.5s idle it fades into the dim track so the bar stops competing
with the conversation. The track column itself stays in layout
regardless, so the viewport never reflows mid-flash (which would
trigger per-item re-measure and a visible jitter).

Implementation kept minimal for stock ink 7:
- gemini-cli's `useAnimatedScrollbar` interpolates RGB colors via
  a theme + per-frame setInterval. The terminal can't render
  smooth fades anyway, so this hook collapses the state to a
  binary `isVisible` flag with a single setTimeout. ~75 LoC.
- `VirtualizedList` calls `flashScrollbar()` from a useLayoutEffect
  keyed on `clampedScrollTop`. The very first commit is skipped
  via a ref so initial mount doesn't paint a flash.
- The render switches the thumb glyph (`█` vs `│`) and `dimColor`
  based on `isVisible && inThumb`. Width stays 1 either way.

Tests (6 new):
- initial mount stays hidden (no spurious mount flash)
- flash → visible, hides after idle timeout, successive flashes
  reset the timer (no premature hide), idleHideMs<=0 disables
  auto-hide for tests that want to assert on the visible state,
  unmount cleans up the pending timer.

Doc updates:
- `docs/design/virtual-viewport/README.md` §1 status, §5 file map,
  §7 PR sequence — V.4 row now scopes only the drag/click-jump
  work (still coord-blocked); animated scrollbar moved out of
  deferred and into shipped.
- PR QwenLM#4146 body — architecture table mentions the auto-hide, new
  files list adds `useAnimatedScrollbar.ts`, test count refreshed
  to 188/188.

Verified: tsc --noEmit -p packages/cli ✓, vitest 188/188 ✓.

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

* fix(cli): VP review round 6 — ESC bug, CI lint, scope-controlled cleanup

Triage of /review feedback from 2026-05-18 + 2026-05-19. Took the
ones that are real and small; declined the ones that are
false-positive / out-of-scope so this PR stops expanding.

Must-fix:
- CI Lint failure: vscode-ide-companion/schemas/settings.schema.json
  was stale after the keyboard-shortcuts description bump. Regenerated
  via `npm run generate:settings-schema`.
- useMouseEvents.ts had `const ESC = '';` (literal empty string after
  the raw 0x1B byte got stripped somewhere in the source pipeline).
  `buffer.indexOf('', 1) === 1` would have degraded garbage skipping
  to a one-byte scan, and the `else { buffer = ''; break }` branch
  could never run. Fixed by switching to the `'\x1b'` text escape and
  doing the same in `mouse.ts` (which had the raw byte, also fragile).
  Comment explains why.

Small wins (one-liners taken from the review batch):
- ScrollableList: rest-spread separates `hasFocus` from the props
  forwarded to VirtualizedList. Latent collision risk; no behaviour
  change today.
- VirtualizedList: `debugLogger.debug` when isReady=false so blank-
  viewport edge cases (tiny terminal / mid-resize race) become
  diagnosable from the debug log instead of looking like a hang.

Real perf (VP-only):
- MainContent: gated the progressive-Static-replay machinery behind
  `!useVirtualScroll`. The render-phase reset still consumes the
  remount-key bump so flag-off toggles mid-session catch up cleanly,
  but `setReplayCount` and the setImmediate chunking effect are now
  skipped for VP users. Saves ~M/CHUNK_SIZE wasted re-renders per
  Ctrl+O / model change on a 1000-turn session.

Belt-and-braces:
- useMouseEvents: added a `process.on('exit')` handler that writes
  the SGR mouse disable seq again. The React cleanup already covers
  normal unmount, but Ctrl+C / SIGTERM / parent kill bypass it and
  the terminal would otherwise stay in button-event-tracking mode
  after qwen exits.

Explicitly declined / deferred (with reasoning logged on the PR):
- requestAnimationFrame wheel throttle: rAF doesn't exist in Node;
  React 19 already batches state updates within a tick, and the
  renderedItems memo bounds the actual work to visible items. Will
  revisit if profiling shows it.
- Stable pending-item IDs (`p-N` keys shifting on completion): the
  observable jitter is at most one frame of estimated-vs-actual
  height delta. Moderate scope (creation-time ID allocation); fits
  better in a focused follow-up than in this PR.

Verified: tsc --noEmit -p packages/cli ✓, vitest 188/188 ✓ across
the full VP suite.

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

* fix(cli): scrollBy bottom uses live end anchor in virtualized list

When keyboard scroll reaches the bottom, scrollBy set isStickingToBottom
but anchored via getAnchorForScrollTop(maxScroll), a fixed {index,offset}
pixel anchor. scrollTo/scrollToEnd instead use {index: last, offset:
SCROLL_TO_ITEM_END}, which recomputes the bottom from live item heights
each render. The fixed anchor did not track the last item growing during
streaming, so scroll-to-bottom via keyboard lagged behind new tokens.
Align scrollBy's bottom branch with the sibling methods.

Reported by wenshao in PR review.

* fix(cli): parse mouse events via ink useInput, not a stdin data listener

useMouseEvents attached its own stdin.on('data', ...) listener. Adding a
'data' listener switches stdin into flowing mode, which drains the buffer
before ink's readable + stdin.read() reader (ink App) can consume it, so
all keyboard input routed through useInput was silently starved while
mouse mode was active.

Parse mouse sequences from ink's existing input pipeline via useInput
instead, so there is only one stdin reader. ink captures a full SGR
sequence (ESC [ < .. M/m) as a single CSI event and delivers it with the
leading ESC stripped, so we re-prepend it before parsing. Non-mouse input
does not match and is ignored; ink still routes input to the app's other
useInput handlers, so keyboard navigation keeps working.

Only SGR mode (1006h, which we enable) is parsed via this path; the legacy
X11 encoding is not recoverable through ink's CSI parser, which is the
encoding modern terminals stop emitting once 1006h is set.

Reported by wenshao in PR review.

* fix(cli): parse only SGR in mouse hook to avoid X11 paste misfire

The useInput-based mouse hook called parseMouseEvent, which also tries the
X11 fallback (parseX11MouseEvent). An X11 prefix (ESC [ M + 3 bytes) can
reach the handler via pasted text — ink emits paste content as input when
no paste listener is registered — and would misfire a spurious mouse event.
Call parseSGRMouseEvent directly so only the SGR encoding we enable (1006h)
is parsed, matching the hook's documented contract.

Reported by wenshao in PR review.

* test(cli): assert SGR mouse parser rejects X11 sequences

Locks in the security property behind the parseMouseEvent ->
parseSGRMouseEvent switch in useMouseEvents: an X11 sequence arriving as
pasted text must not misfire a mouse event. Asserts a well-formed X11
sequence is a valid X11 event yet returns null from parseSGRMouseEvent, so
a future revert to parseMouseEvent fails this test.

Reported by wenshao in PR review.

* test(cli): add VP scroll coverage + eslint-disable for useBatchedScroll

Cover keyboard scroll commands (Shift+Up/Down, PageUp/Down, Ctrl+Home/End),
scrollBy/scrollTo imperative API (positive/negative/overflow/clamp), and
auto-scroll-during-streaming state machine (stick-to-bottom, disengage on
user scroll, re-engage on scrollToEnd). Add missing eslint-disable-next-line
for intentionally dep-free useLayoutEffect in useBatchedScroll.

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

* chore(cli): remove trailing whitespace in useBatchedScroll

The eslint-disable-next-line comment was removed by eslint --fix as an
unused directive (exhaustive-deps does not flag a useLayoutEffect with
no dependency array). Clean up the residual blank line.

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

---------

Co-authored-by: 秦奇 <[email protected]>
Co-authored-by: Qwen-Coder <[email protected]>

…M#4414)

PR QwenLM#4064 introduced ~/.qwen/file-history/{sessionId}/ for /rewind but had
no cross-session cleanup — directories accumulated indefinitely. This adds
a generic background housekeeping framework with file-history cleanup as
its first user.

- 30-day mtime sweep, configurable via general.cleanupPeriodDays
- 10-min startup delay (1-min catch-up if last run >7d ago)
- 24h recurring cadence, idle-gated (defers if user typed in last 1 min)
- O_EXCL lockfile + marker mtime throttle (multi-process safe)
- Current session whitelisted via lazy config.getSessionId() — defends
  against long-idle active sessions and /clear minting a new session
- Negative cleanupPeriodDays values clamp to 1h minimum (defends against
  schema-bypass: a future cutoff would otherwise sweep everything)
- Zero new prod dependencies; ~70 lines of self-written O_EXCL throttle
  primitive in lieu of proper-lockfile (which pulls graceful-fs and
  monkey-patches every fs method on first require)
- All setTimeout(...).unref() — never blocks process exit

Closes QwenLM#4173.

🤖 Generated with [Qwen Code](https://github.com/QwenLM/qwen-code)

…king (QwenLM#4680)

* fix(core): loosen auto-mode classifier timeouts, disable stage-2 thinking

The AUTO-mode classifier fails closed on timeout — a timed-out judge call
blocks the action as "unavailable". The tight 3s/10s stage budgets turned
transient slowness (slow network, large transcript, model queueing) into
spurious blocks of otherwise-valid actions. Raise them to 10s/30s so a
slow-but-healthy call is not treated as a hard block.

Also disable thinking in stage 2 (previously the only stage with
includeThoughts: true). This is a latency-sensitive permission gate the
user is actively waiting on; allocating a reasoning budget made the review
path slower and more expensive, which directly worsened the fail-closed
timeout. The model still records its reasoning in the structured
`thinking` output field — it just no longer gets an allocated budget.

Closes QwenLM#4676

Co-authored-by: Qwen-Coder <[email protected]>

* docs(core): trim verbose comments in auto-mode classifier

Condense the three comments touched by this change (module docstring
stage-2 note, timeout-budget rationale, stage-2 thinkingConfig) while
keeping the essential "why". No logic changes.

Co-authored-by: Qwen-Coder <[email protected]>

---------

Co-authored-by: Qwen-Coder <[email protected]>
Co-authored-by: Qwen-Coder <[email protected]>

…rt 1) (QwenLM#4439)

* fix(core): coerce hostile-provider usage token counts (QwenLM#4350 part 1)

Hostile providers (broken upstream, OpenAI-compat proxy returning
null/NaN, misconfigured override) can emit non-finite or negative
values for `usageMetadata.{prompt,candidates,cached,total}TokenCount`.
Captured unguarded in `processStreamResponse`, these poison the
compaction gate arithmetic:

- `lastPromptTokenCount + NaN >= hard` is always false → hard-rescue
  is silently disabled, eventually OOMing the V8 heap.
- `Infinity >= hard` is always true → hard-rescue fires every send.

Route the four API capture sites through a `coerceUsageCount` helper
that maps unknown / non-finite / negative to 0. `Number.isFinite(-1)`
is true, so an explicit `>= 0` is needed in addition to `isFinite`.

Part 1 of the hostile-provider hardening from QwenLM#4350. The companion
`computeThresholds` guard depends on the un-merged three-tier ladder
in QwenLM#4345 and is deferred until that lands.

Covered by parametrized tests in `geminiChat.test.ts` over NaN,
±Infinity, negative, null, undefined, and string inputs, plus a
fallback test asserting a hostile `promptTokenCount` falls through
to a coerced `totalTokenCount`.

* docs(core): narrow coerceUsageCount JSDoc to fields it actually coerces

The JSDoc for coerceUsageCount enumerated
{prompt,candidates,cached,total}TokenCount as the protected fields, but
candidatesTokenCount is never passed through coerceUsageCount in this PR --
only promptTokenCount, totalTokenCount, and cachedContentTokenCount are.
The original wording created a misleading impression of coverage.

Rewrite the JSDoc to (a) list exactly the three fields the function is
called on, (b) state explicitly that candidatesTokenCount is intentionally
left raw because it does not feed the compaction gate, and (c) call out
that candidatesTokenCount still flows unguarded into OTel spans via
loggingContentGenerator -- to be tightened separately.

No behavior change. Resolves the geminiChat.ts:109 review thread on this
PR.

* fix(core): address review feedback on coerceUsageCount

- Fix JSDoc: 'This PR' → 'This function' for long-term documentation
- Add optional field name parameter to coerceUsageCount for debug logging
- Log hostile values via debugLogger.warn when coercion fires
- Coerce candidatesTokenCount (was missing despite JSDoc mentioning it)
- Build sanitized usageMetadata with coerced values for recordAssistantTurn
  so JSONL persistence and --resume don't re-poison the compaction gate

* fix(core): reuse coerced usage values in recordAssistantTurn

The streaming loop was calling coerceUsageCount twice for the same
usageMetadata: once for telemetry updates and again inline when building
the tokens object for recordAssistantTurn. This left candidatesTokenCount
declared but unused in the first block (TS6133).

Now stashes all four coerced values in a coercedUsage object during the
streaming loop, then spreads it into the tokens object when recording.
This eliminates the duplicate coercion calls and uses candidatesTokenCount
consistently.

Resolves review thread: PRRT_kwDOPB-92c6Ed4WT

* test(core): assert debugLogger.warn for hostile usage counts

Address PR QwenLM#4439 review thread on geminiChat.test.ts: the hostile-provider
parametrized test asserted token-count coercion behaviour but never verified
the operator-facing diagnostic emitted by `coerceUsageCount`. A future
refactor could remove or misformat the warning with no test failure.

- Mock `createDebugLogger` via vi.hoisted/vi.mock so the module-level
  `debugLogger` in geminiChat.ts routes warnings to a spy.
- Inside the existing it.each, assert that hostile defined values
  (NaN/Infinity/-Infinity/negative/string) emit warns mentioning both
  `hostile promptTokenCount` and `hostile totalTokenCount`, and that the
  stringified bad value is included so logs are actionable.
- Add a negative assertion that the field-omitted cases (null, undefined)
  do NOT trigger any warn — verifying the `value != null` guard.

…ell subprocesses (QwenLM#4649)

* feat(core): inject context env vars (session/agent/prompt ID) into shell subprocesses

When SubAgents execute SQL or Python scripts via Bash tool, the scripts
have no way to know their execution context. This adds automatic injection
of QWEN_CODE_SESSION_ID, QWEN_CODE_AGENT_ID, and QWEN_CODE_PROMPT_ID
into all shell subprocess environments, enabling downstream scripts to
perform trace correlation, audit logging, and business context attribution.

Closes QwenLM#4645

* fix(core): use module-level flag to guard session env claim

Prevents nested qwen-code processes from inheriting the parent's
session ID. The previous `if (!process.env[...])` check would keep
the parent's value; a module-level flag ensures each process claims
its own session ID on first Config construction.

* fix(test): use bracket notation for index signature properties

CI tsc --build enforces noPropertyAccessFromIndexSignature; use
env['KEY'] instead of env.KEY to satisfy strict type checking.

* fix(core): guard process.env assignment for mocked process environments

Some test suites mock node:process without providing env, causing
TypeError when Config constructor assigns QWEN_CODE_SESSION_ID.
Add defensive check before env writes.

* test(config): add coverage for sessionEnvClaimed guard

Addresses reviewer feedback (wenshao) requesting test coverage for the
module-level `sessionEnvClaimed` guard in Config constructor.

Tests verify:
1. First Config instance sets process.env['QWEN_CODE_SESSION_ID']
2. Subsequent Config instances do not overwrite the env var

* test(config): add startNewSession env var test and clarify comment

- Add test verifying startNewSession updates process.env to new session ID
- Add comment explaining why startNewSession bypasses sessionEnvClaimed guard
  (only callable on the canonical Config instance that already claimed)

* feat(core): add auto-mode denial observability

* fix(core): prune unused auto denial reason

* fix(core): scope auto fallback counter resets

* test(core): cover denied fallback cancellation

* fix(core): tighten auto-mode denial fallback tracking

* test(core): cover permission denied hook events

* fix(core): preserve auto fallback reasons

* test(core): cover auto mode denial helpers

* fix(core): reset auto fallback after hook approval

* fix(core): log auto fallback reset recovery

* fix(core): clarify auto fallback reset logging

* test(core): assert auto fallback reset logging

QwenLM#4654)

* feat(core): auto-dump memory diagnostics to disk on pressure detection

When the MemoryPressureMonitor (QwenLM#4403) detects hard or critical pressure,
write a lightweight diagnostics JSON to .qwen/<project>/diagnostics/ before
running cleanup. The file survives even if a subsequent operation triggers
OOM, giving maintainers actionable data from bug reports without requiring
the user to manually run /doctor memory after a crash.

Design follows Claude Code's heapDumpService approach: write the cheap JSON
first (small write, won't OOM), heavy snapshot second. Diagnostics include
process memory stats, V8 heap stats, session history size, and an actionable
suggestion for the user.

Per-session limits: max 3 dumps, 30s cooldown between dumps.

Closes QwenLM#4651

* ci: retrigger CI after Windows flaky failure

* test(core): use path.join in memoryDiagnosticsDumper test for cross-platform

The assertion hard-coded POSIX separators ('/tmp/test-project/diagnostics/'),
which fails on Windows where path.join produces backslashes. Build the
expected substring with path.join + path.sep so it matches the dumper's
actual output on every platform.

* fix(core): two-phase memory diagnostics write to survive OOM

Two critical issues from review:

1. The async collectMemoryDiagnostics() runs before writeFileSync, but it
   spawns a `ps` subprocess and reads /proc — fork() under critical memory
   pressure can fail or be OOM-killed, leaving no file on disk despite the
   "cheap write first" design comment.

2. dumpCount and lastDumpTime were updated after the await, so concurrent
   dumps (e.g. hard→critical escalation) would both pass the cap/cooldown
   guards and overwrite each other.

Fix:

- Reserve the dump slot synchronously (++dumpCount, lastDumpTime) before
  any await, so concurrent calls correctly hit the cap.
- Phase 1: synchronously write a minimal JSON (process.memoryUsage +
  v8.getHeapStatistics, no fork/exec) with collectionComplete=false.
  Because async functions execute synchronously up to the first await,
  this is guaranteed on disk before the caller's next statement runs.
- Phase 2: enrich with full diagnostics asynchronously and overwrite the
  file with collectionComplete=true. If Phase 2 crashes, the minimal
  Phase 1 file still survives for debugging.

Tests updated for the two-phase write and gain two new cases covering the
sync-Phase-1 guarantee and the synchronous slot reservation.

* fix(core): point memory diagnostics suggestion at /compress (the actual command)

The suggestion text told users to run /compact, which does not exist in
this repository — the actual command is /compress (see compressCommand.ts).
Pointing users at a nonexistent slash command in a diagnostics report
makes the suggestion unactionable.

…NL (closes QwenLM#3681, QwenLM#4095 Phase 2) (QwenLM#4333)

* feat(core): add atomicWriteFileSync + forceMode option

Sync mirror of atomicWriteFile for code paths that can't await
(settings persistence on exit, sync config writers). Same semantics:
symlink chain resolution, permission preservation, fsync via flush:true,
EPERM/EACCES rename retry, EXDEV fallback to direct write.

Add forceMode option on AtomicWriteFileOptions — when true, ignore the
existing target's permission bits and apply options.mode regardless.
Needed for credential files that must heal historically over-permissive
files (e.g. a 0o644 token restored from backup must be forced to 0o600).
Honored by both async and sync paths. Default false preserves existing
behavior.

Reuses Atomics.wait for true blocking sleep in renameWithRetrySync —
no busy-wait, no extra dep.

Refs: QwenLM#4095 Phase 2

* refactor(core): migrate credential writes to atomicWriteFile (QwenLM#4095 Tier 1)

Route all OAuth credential persistence through atomicWriteFile with
forceMode: true, so a process crash mid-write cannot leave the user
with a half-written token file, and historically over-permissive files
(e.g. 0o644 from a manual restore) are healed to 0o600 on the next
write.

- oauth-token-storage.ts: setCredentials, deleteCredentials
- file-token-storage.ts: saveTokens (encrypted MCP token storage)
- qwenOAuth2.ts: cacheQwenCredentials (also fixes missing mode — was
  inheriting 0o644 from umask, now forced to 0o600)
- sharedTokenManager.ts: saveCredentialsToFile — drops ~15 lines of
  hand-rolled tmp + rename in favor of the shared helper

Lock-file writes using flag: 'wx' (sharedTokenManager.ts:720) are
intentionally left untouched — they rely on exclusive-create semantics
that atomic write does not preserve.

Tests updated to mock atomicWriteFile instead of fs.writeFile.

Refs: QwenLM#4095 Phase 2

* refactor(core): migrate memory state writes to atomicWriteFile (QwenLM#4095 Tier 2)

Route all auto-memory state persistence through atomicWriteFile so a
process crash during a dream/extract/forget cycle cannot corrupt the
metadata sidecar, extraction cursor, or topic body files.

Touched: manager (writeDreamMetadata), extract (writeExtractCursor +
bumpMetadata), indexer (rebuild), dream (bumpDreamMetadata), forget
(bumpMetadata + topic body rewrite).

manager.ts:362 acquireDreamLock uses flag: 'wx' for exclusive create —
left untouched, atomic write does not preserve that semantic.

Uses atomicWriteFile (not atomicWriteJSON) to preserve the trailing
newline these files have always had.

Refs: QwenLM#4095 Phase 2

* refactor: migrate config + logger + state writes to atomic helpers (QwenLM#4095 Tier 3a)

Route the remaining state-file write paths through atomic helpers so a
crash mid-write cannot corrupt config, log, or session-scoped state:

- trustedFolders.ts (sync): atomicWriteFileSync — sole path that flips
  workspace trust, must not half-write
- logger.ts (4 sites): atomicWriteFile — full-file JSON rewrites for
  logs.json and per-checkpoint files
- tipHistory, installationManager, projectSummary, todoWrite,
  trustedHooks: bonus sites with the same shape (state JSON written
  multiple times per session)

todoWrite is on the hot path — writes every time the todo list mutates
— so the added rename + fsync cost is measurable (a few ms per write
on SSD). Trade-off accepted to avoid a half-written todos file
silently breaking the next session's resume.

Export atomicWriteFile / atomicWriteFileSync from the core public
index so CLI-side callers (trustedFolders, tipHistory) can reach them.

Tests updated:
- logger.test.ts uses vi.importActual to re-export the real helper and
  override per-test via vi.mocked(atomicWriteFile).mockRejectedValueOnce
- trustedFolders.test.ts and todoWrite.test.ts mock the helper directly

Refs: QwenLM#4095 Phase 2

* fix(core): flush JSONL appends to disk (QwenLM#4095 Tier 3b, closes QwenLM#3681)

QwenLM#3656 fixed the read side of glued '}{' JSONL records — when a process
was killed mid-appendFile, the trailing '\n' was lost and the next
record was concatenated. The write side was left for a follow-up
(QwenLM#3681).

This adds flush:true (fsync) to every per-line append:
- jsonl-utils.ts writeLine / writeLineSync (session transcripts,
  auto-titles, prompt history)
- debugLogger.ts appendFile (per-session debug log)

jsonl-utils.ts write() (full-file replace) now goes through
atomicWriteFileSync so a crash during overwrite cannot corrupt the
session transcript either.

Trade-off: fsync on every append adds disk-sync latency (single-digit
ms on SSD, more on spinning disk / network FS). Acceptable for a few
writes per turn; the alternative is silently losing the last record
of every interrupted session, which QwenLM#3681 explicitly flagged.

Refs: QwenLM#4095 Phase 2 Tier 3b
Closes: QwenLM#3681

* refactor(core): migrate extension config + LSP edit to atomic write

Catch up two sites where Claude Code's equivalent path is atomic but
qwen-code's isn't (verified against
/Users/jinye.djy/Projects/claude-code on 2026-05-19):

- extension/extensionManager.ts:533, :1073 — enablement config and
  install metadata writes. Claude Code's plugin install-counts and zip
  cache use atomic temp+rename via writeFileSyncAndFlush_DEPRECATED.
- lsp/NativeLspService.ts:1351 — applying an LSP edit to a user file.
  Claude Code's FileWriteTool/FileEditTool both route through atomic
  writeTextContent → writeFileSyncAndFlush_DEPRECATED. A bare
  writeFileSync here could half-write the user's source file if the
  process is killed during an LSP-driven rename or quick-fix.

Also clean up stale fs.rename mock setups in sharedTokenManager.test.ts
that became no-ops after Tier 1 migration (rename is no longer called
by saveCredentialsToFile). The fs.writeFile mocks stay because the
wx-flag lock path still uses them.

Refs: QwenLM#4095 Phase 2

* chore: cosmetic cleanups from PR review

- packages/core/src/index.ts: move atomicFileWrite export to its
  alphabetical position (before browser.js)
- tipHistory.ts: add forceMode: true to atomicWriteFileSync for
  consistency with other 0o600 sites — heals legacy 0o644 files even
  though tips are non-critical

Refs: QwenLM#4095 Phase 2

* fix(core): address Codex review findings on Phase 2 PR

Three issues caught by post-merge Codex review of the QwenLM#4095 Phase 2
branch — none had user-visible symptoms yet but all were latent bugs.

1. atomicFileWrite: forceMode without mode silently downgraded perms
   `if (!options?.forceMode)` skipped the existing-mode stat whenever
   forceMode was true, regardless of whether `mode` was also supplied.
   Calling `atomicWriteFile(p, data, { forceMode: true })` (no mode) on
   an existing 0o600 file produced 0o644 (umask default) instead of
   preserving 0o600. Tightened the guard to also require `options.mode`
   to be defined; mirrored fix in atomicWriteFileSync. Added two
   regression tests (async + sync) that assert mode preservation.

2. logger.test.ts: vi.resetAllMocks() blanked the atomicWriteFile shim
   The vi.fn(actual.atomicWriteFile) factory implementation gets reset
   to a no-op by `vi.resetAllMocks()` in beforeEach, which would make
   `logger.initialize()` silently skip creating logs.json on disk.
   Tests passed by coincidence (file pre-existence from prior runs).
   Captured the real implementation at module load and re-attach it via
   `mockImplementation` after each reset.

3. NativeLspService.applyTextEdits: atomic write bypassed file unwritability
   The read catch swallowed every error and treated it as "new empty
   file". With atomic write (tmp + rename), an unreadable target on a
   writable parent could be replaced with edits applied to an empty
   buffer — the old fs.writeFileSync would have errored on the target
   permission. Now only ENOENT is treated as new-file; other read
   errors (EACCES, EISDIR, etc.) propagate.

Refs: QwenLM#4095 Phase 2

* fix(lsp): refuse LSP edits to chmod 0444 files (Codex round 2)

The previous fix only handled "read failed → propagate the error".
Codex round 2 caught the remaining gap: a file that's readable but
chmod 0444 (read-only) would still be replaced by the atomic rename,
because rename only needs parent-directory write access.

Add an explicit fs.accessSync(W_OK) check before the atomic write.
ENOENT is allowed through so LSP can still create new files via edits.

Refs: QwenLM#4095 Phase 2

* fix(core): drop withTimeout around atomic credential write (Codex round 3)

`saveCredentialsToFile` wrapped `atomicWriteFile` in `withTimeout(5000)`.
If the call hits the 5s budget (e.g. slow NFS home, network-backed
storage, fsync added by Phase 2), withTimeout rejects but the
atomicWriteFile internal write+rename keeps running unobserved:

  1. withTimeout rejects → saveCredentialsToFile throws
  2. performTokenRefresh `finally` releases the refresh lock
  3. Another process acquires the lock and writes newer credentials
  4. The original atomicWriteFile finally completes its rename and
     overwrites the newer credentials — silent token rollback

Pre-migration the code awaited the tmp write and the rename in two
separate withTimeout calls; a timed-out tmp write never reached the
rename so there was no race against the target file. The migration
collapsed both into one inseparable atomicWriteFile, which made the
timeout actively unsafe (the work cannot be cancelled after the
timeout fires — fs.rename is not abortable).

Atomic write is durable by design — accept the I/O latency. The
mkdir and stat timeouts are kept (idempotent and read-only
respectively, no corruption risk on late completion).

Refs: QwenLM#4095 Phase 2

* test(core): add rename-retry + EXDEV-fallback coverage (QwenLM#4333 review)

Address PR review suggestions from wenshao (via qwen-latest /review):
neither renameWithRetry/Sync nor the EXDEV cross-device fallback had
direct test coverage. Both paths are critical (Windows AV contention,
Docker tmpfs /tmp) and a regression would degrade silently.

Vitest can't spy on ESM exports of `node:fs` (`Cannot redefine property:
renameSync`), so add narrow internal test seams instead:
- renameWithRetry / renameWithRetrySync take an optional `_renameImpl`
  parameter, defaulting to fs.rename / fs.renameSync.
- atomicWriteFile / atomicWriteFileSync take an optional `_testFs`
  parameter with `rename` and `writeFile` overrides, forwarded to the
  retry helper and used in the EXDEV fallback branch.

The seams are underscore-prefixed and JSDoc-tagged as "Internal test
seam — production callers never pass this", which keeps the public API
clean while making the behavior testable.

New coverage (+9 tests, 36 → 45):
- renameWithRetry: retry-EPERM-then-succeed, give-up after retries,
  no-retry on non-retryable (ENOSPC)
- renameWithRetrySync: same 3 patterns (EACCES, EPERM exhausted, EINVAL)
- EXDEV fallback: async direct write + tmp cleanup, sync ditto,
  non-EXDEV failure propagates without fallback (rejects EIO + tmp cleanup)

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(test): update telemetry sdk.test.ts appendFile assertions for flush:true

CI failure on all 3 OSes (macos / ubuntu / windows): sdk.test.ts asserted
`fs.appendFile` was called with `'utf8'` as the 3rd positional argument,
but commit b7badc7 (QwenLM#4095 Tier 3b — JSONL fsync) changed the
`debugLogger.ts` appendFile call from string-form to options-form
`{ encoding: 'utf8', flush: true }` to enable per-line fsync. Update the
3 assertions in the telemetry diagnostics test to match the new shape.

No behavior change — debugLogger still flushes per append; only the
assertion in this previously-unrelated suite needed updating.

Refs: QwenLM#4333, QwenLM#4095 Phase 2 Tier 3b

* test: cover LSP error branches + sync EXDEV cleanup + JSONL writes (QwenLM#4333 review)

Address three review suggestions from wenshao (via qwen-latest /review),
each pointing at a real coverage gap introduced by this PR:

1. NativeLspService.applyTextEdits error branches (round-2 LSP fix):
   the ENOENT-only read guard and the fs.accessSync(W_OK) refusal had
   no automated coverage. Added 3 tests accessing applyTextEdits via a
   typed cast (the method is private; making it public for one
   verification inflates API surface). Tests use chmod 0000 / chmod 0444
   reproducers and assert (a) read failure propagates EACCES without
   silently overwriting with empty content, (b) W_OK rejects with
   EACCES/EPERM before the atomic rename touches the target,
   (c) nonexistent files are still accepted so LSP can create via edits.

2. atomicWriteFileSync non-EXDEV rename failure cleanup: the async
   counterpart had an explicit EIO-rename test asserting tmp cleanup;
   the sync variant did not. Added the mirror — injects a sync rename
   throwing EIO via the existing _testFs seam and asserts
   `readdirSync(tmpDir).length === 0`.

3. jsonl-utils writeLine / writeLineSync / write smoke tests: the three
   write paths are the core fix for QwenLM#3681 (the PR's headline goal) but
   downstream callers (chatRecordingService, sessionService) mock them
   entirely. Without direct unit tests, a regression that dropped
   `flush: true` or reverted `write()` to bare writeFileSync would go
   undetected. Added 3 real-fs roundtrip tests.

Test count delta:
- NativeLspService.test.ts: 15 → 18
- atomicFileWrite.test.ts: 45 → 46
- jsonl-utils.test.ts: 22 → 25

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core,test): wrap atomic write errors + guard root user + cover save failure (QwenLM#4333 review)

Three review fold-ins from wenshao (via qwen-latest /review):

1. atomicFileWrite: error messages reference the random `.tmp.<hex>`
   path, not the logical target — many callers (memory subsystem,
   extension manager) don't wrap the error, making debug logs unhelpful.
   Add `annotateWriteError(error, targetPath)` that mutates the error
   message in-place to prefix `atomicWriteFile("<targetPath>"): ` while
   preserving `code` / `errno` / `syscall` / `stack` / the prototype
   chain so downstream `err.code === 'ENOENT'` checks and `instanceof`
   narrowing keep working. Applied to both async and sync variants;
   only the final propagated throw (not the EXDEV fallback path) is
   annotated.

2. NativeLspService.test.ts: the chmod 0444 and chmod 0000 tests rely
   on `accessSync(W_OK)` and `readFileSync` failing — but on POSIX with
   UID 0 (root, including most Docker CI runners), permission bits are
   bypassed and `accessSync` always succeeds. The tests would silently
   pass even with the W_OK guard removed entirely. Add
   `process.getuid?.() === 0` to the skip guard on both tests.

3. sharedTokenManager.test.ts: the catch block in saveCredentialsToFile
   that maps disk-full / permission-denied to
   `TokenManagerError(FILE_ACCESS_ERROR)` was never exercised — every
   prior test mocked atomicWriteFile as always-successful. Added a
   regression test that rejects atomicWriteFile with ENOSPC and asserts
   the wrapped TokenManagerError surfaces with the right type and
   carries the original message.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core,lsp): annotate sync errors correctly + cover EXDEV fallback + async LSP IO (QwenLM#4333 review)

Three review fold-ins from wenshao (via qwen-latest /review), all real
correctness/consistency issues:

1. annotateWriteError hardcoded "atomicWriteFile" prefix regardless of
   caller. Sync write failures produced misleading
   `atomicWriteFile("/path"):` prefixes in incident logs. Add optional
   `fnName` parameter (defaults to async name) and have sync call sites
   pass "atomicWriteFileSync".

2. EXDEV fallback path in both async and sync variants did NOT route
   the inner writeFileImpl/tryChmod failures through annotateWriteError.
   On a cross-device write that subsequently hit ENOSPC, the propagated
   error had a bare syscall message without the target-path prefix —
   breaking the function's documented error-shape contract. Wrap both
   fallback branches in try/catch + annotate.

3. NativeLspService.applyTextEdits is declared `async` and all callers
   `await` it, but the round-2 fix mixed in sync IO: readFileSync,
   accessSync, atomicWriteFileSync. The sync helper's renameWithRetrySync
   blocks the event loop up to ~350ms under Atomics.wait EPERM backoff
   — particularly bad for LSP workspace edits that loop over many files.
   Switch to async throughout: fsp.readFile, fsp.access, atomicWriteFile.
   Behavior preserved (same ENOENT-vs-other distinction, same W_OK gate).
   Existing tests pass unchanged (they already use the async typed-cast
   entry point).

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* test(core): cover EXDEV-fallback-write-failure annotation path (QwenLM#4333 review)

The previous fold-in added try/catch around the EXDEV fallback write in
both async and sync variants, routing fallback failures through
`annotateWriteError(err, target, fnName)`. The sync-specific
`fnName='atomicWriteFileSync'` differentiation is ONLY exercised on
that path, so a regression that dropped or misapplied the annotation
on sync would otherwise go undetected.

Two new tests inject both a failing rename (EXDEV) and a failing
writeFile (ENOSPC) via the `_testFs` seam, then assert (a) the
original `code === 'ENOSPC'` propagates intact, and (b) the message
matches `/atomicWriteFile(Sync)?\(<target>\):.*ENOSPC/` — verifying
target-path prefix AND correct fn-name differentiation.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core): annotate guard, drop debug fsync, add noFollow for creds (QwenLM#4333 review)

Four review fold-ins from wenshao (via qwen-latest /review):

1. annotateWriteError guard `!error.message.includes(targetPath)` was
   always false in production. tmpPath is `${targetPath}.<hex>.tmp`, so
   any real fs error like `ENOSPC ... '/path/creds.json.a1b2c3.tmp'`
   *contains* targetPath as substring → guard returned false → the
   annotation prefix was silently skipped on every real failure path.
   Tests passed only because mock errors used bare `new Error('ENOSPC')`
   messages without paths. Change guard to idempotency check on our
   own prefix (`startsWith(`${fnName}(`)`), and update the two existing
   EXDEV-fallback tests to use realistic path-embedding fs messages.

2. debugLogger.appendFile dropped `flush: true`. 1050+ call sites,
   default-enabled, fire-and-forget — per-line fsync creates sustained
   I/O pressure / SSD wear with no user benefit (debug logs are
   best-effort, the module already tracks `hasWriteFailure` for the
   degraded-mode UI). Kept the kernel page-cache flush; revert
   debugLogger.test.ts and telemetry/sdk.test.ts assertions back to
   plain `'utf8'`. The QwenLM#3681 closure target is jsonl-utils writeLine,
   not debug logs.

3. Symlink security regression: the old
   `fs.writeFile(tmp) + fs.rename(tmp, filePath)` pattern atomically
   *replaced* a pre-placed symlink at `filePath`. atomicWriteFile's
   default `resolveSymlinkChain(filePath)` follows the link and writes
   through, redirecting tokens to wherever the link points (real
   concern on shared hosts with weaker-than-expected dir perms). Add
   `noFollow?: boolean` option that skips chain resolution; apply
   `noFollow: true` to all 4 credential write sites
   (oauth-token-storage [2 sites], file-token-storage, sharedTokenManager)
   to match the pre-migration replace-symlink semantics.

4a. Test seam `_testFs` was a 4th positional arg → considered moving
    into options. Punted: positional with underscore + JSDoc is
    materially the same surface as options field with underscore,
    and the only realistic collision (future production option as
    5th arg) is bounded by review.
4b. Sync/async code duplication (~110 lines mirror) → DECLINED.
    Refactoring to a sync/async-polymorphic helper introduces a new
    abstraction layer with worse type ergonomics; the duplication is
    mechanical and lined up for easy diffing. Tracked as Phase 2.5
    candidate if divergence actually accumulates.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core): EXDEV fallback honors noFollow + fix test correctness (QwenLM#4333 review)

Three review fold-ins from wenshao (via qwen-latest /review):

1. **[CRITICAL]** EXDEV fallback path silently bypassed `noFollow`
   protection. The fallback `writeFileImpl(targetPath, ...)` is
   `fs.writeFile` / `fs.writeFileSync`, both of which follow symlinks.
   When a credential write site set `noFollow: true` and rename threw
   EXDEV (realistic on Docker OverlayFS / union mounts), the fallback
   would write credentials *through* a pre-placed symlink to an
   attacker-controlled target — the exact attack noFollow was meant
   to prevent. Fix: when `noFollow` is set, the EXDEV fallback now
   `unlink`s any existing entry, then opens with `O_WRONLY|O_CREAT|O_EXCL`
   to refuse to write through a symlink that races back. Applied to
   both async and sync variants.

2. **Test correctness**: the two EXDEV-fallback-write-failure tests
   added in the previous round had a bug — the `failingWrite` mock
   threw on every call, including the first call which is the
   tmp-file write. The tmp write failed → outer catch caught
   ENOSPC → EXDEV check returned false (code is ENOSPC) → fell
   through to the outer annotateWriteError. The inner
   EXDEV-fallback annotation was never exercised. The assertions
   passed only because annotateWriteError produces the same format
   in both catch blocks. Fix: selective-failure mock that succeeds
   on the first call (tmp write) and fails on the second (fallback
   write), genuinely reaching the EXDEV branch.

3. **Behavioral noFollow tests**: previously `noFollow: true` was
   only verified at the "option is passed to mock" level. Added 4
   real-fs tests (async + sync × happy-path + EXDEV-fallback) that
   pre-place a symlink, call atomicWriteFile with noFollow, and
   assert: (a) the symlink is replaced by a regular file, (b) the
   new file holds the new data, (c) the real file behind the
   symlink is untouched. A regression flipping the noFollow ternary
   or skipping the noFollow-aware EXDEV fallback now fails directly.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core): close TOCTOU window on noFollow EXDEV fallback + cover ENOENT path

Two PR QwenLM#4333 round-7 review items folded in:

1. TOCTOU race: the path-based `tryChmod(targetPath)` after the EXDEV
   noFollow branch ran AFTER `fd.close()` released the inode reference.
   Between close and chmod, an attacker with parent-directory write access
   could replace the regular file with a symlink, redirecting the
   `chmod 0o600` onto an attacker-chosen target — silently defeating the
   `noFollow` protection that the `unlink + O_EXCL` pattern was added to
   provide. Fix: switch to `fd.chmod`/`fchmodSync` on the open fd before
   close (operates on the inode, immune to symlink swap), and skip the
   path-based chmod for the noFollow branch (path-based chmod remains
   for the non-noFollow direct-write branch, where following symlinks
   was already in scope).

2. Missing test coverage: all 4 existing noFollow EXDEV tests pre-place
   a symlink at the target, so `unlink(targetPath)` always succeeded —
   the ENOENT-swallow branch (first-write scenarios, e.g. initial
   credential provisioning on a cross-device mount) had no coverage.
   Added 2 tests (async + sync) verifying the fallback creates a new
   file with the requested mode when the target never existed.

Test results: 54/54 atomicFileWrite tests pass (was 52). Async + sync
parity preserved.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* test(core): skip new noFollow EXDEV-fallback tests on Windows (NTFS perm bits)

The two new-file noFollow EXDEV tests added in 7c47664 assert the file
ends up at 0o600. NTFS reports 0o666 for any file that isn't read-only
regardless of chmod/fchmod, so the assertion fails on Windows runners
(`AssertionError: expected 438 to be 384`). Match the existing
`it.skipIf(process.platform === 'win32')` pattern already used by all
mode-asserting tests in this file. Linux/macOS coverage of the
ENOENT-swallow + new-file path is unchanged.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core): narrow fchmod catch + verify mechanism on noFollow EXDEV

Two PR QwenLM#4333 round-8 review items:

1. The `catch {}` around `fd.chmod(desiredMode)` (and `fchmodSync`) was
   intended to tolerate filesystems without POSIX permissions (FAT/exFAT)
   but silently swallowed every error code: EPERM under a hardened
   sandbox, EIO on flaky NFS/CIFS, EROFS if the filesystem is remounted
   read-only mid-write. Because this path operates on the *final
   credential file* — not a tmp file — a silent fchmod failure leaves
   the target at the umask-masked open() mode with no diagnostic trail.
   Narrowed the catch to ENOSYS/ENOTSUP so the FAT/exFAT case still
   tolerates failure but security-relevant errors propagate.

2. The round-7 noFollow-EXDEV-new-file tests asserted the final mode
   (0o600) but didn't verify the *mechanism*. Under typical umask 0o022,
   `open(O_EXCL, 0o600)` already creates the file at 0o600, so a
   regression that swapped `fd.chmod()` back to a path-based
   `tryChmod(targetPath)` (the pre-fix TOCTOU-vulnerable form) would
   leave the mode assertion passing — defeating the round-7 fix
   undetected. Added a `chmod` test seam to `atomicWriteFile` /
   atomicWriteFileSync` (mirroring the existing `rename` / `writeFile`
   seams) and asserted that path-based chmod is never invoked against
   the credential target on this code path.

54/54 atomicFileWrite tests pass.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core): clean up O_EXCL orphan on noFollow EXDEV fchmod failure + cover both branches

Two PR QwenLM#4333 round-9 review items, one critical correctness bug
introduced by the round-8 catch narrowing:

1. **[Critical]** When `fd.chmod(desiredMode)` (or `fchmodSync`) on the
   noFollow EXDEV fallback throws a propagating error (EPERM under
   seccomp, EIO on flaky NFS, EROFS after a remount), the file created
   by `open(targetPath, O_CREAT|O_EXCL)` remained on disk after the
   error rethrew. Every subsequent credential write retry then hit
   EEXIST from O_EXCL and failed permanently — turning a transient
   sandbox EPERM into a permanent OAuth-refresh deadlock that requires
   manual file removal. All three credential write sites
   (`sharedTokenManager`, `oauth-token-storage` save+delete,
   `file-token-storage`) hit this code path. Fix: a `writeOk` flag plus
   nested try/catch — fd is closed in the inner finally, then if write/
   sync/fchmod failed, the orphan is unlinked best-effort before the
   error rethrows.

2. The fchmod catch-narrowing (the headline behavior of round 8) had
   zero test coverage on either branch — `_testFs` exposed `rename`,
   `writeFile`, and path-based `chmod`, but no hook for the open-fd
   `fchmod`. A one-line revert from the narrowed catch back to
   `catch {}` would pass every existing test. Added `fchmod` field to
   `_testFs` (async + sync) and four tests:
   - ENOSYS swallowed → write succeeds (FAT/exFAT happy path)
   - EPERM propagates AND `targetPath` is absent (regression-tests #1)
   for both async and sync.

58/58 atomicFileWrite tests pass (was 54). Async + sync parity preserved.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core): annotate symlink errors + cover ENOTSUP/EACCES + log orphan unlink

Four PR QwenLM#4333 round-10 review items, all small bounded fixes:

1. Parameterized the FAT/exFAT fchmod-swallow tests over both ENOSYS
   (Linux) and ENOTSUP (macOS). Round-9 only covered ENOSYS — a
   one-token regression dropping `ENOTSUP` from the catch condition
   would have passed every existing test.

2. The orphan-unlink catch block was empty (`/* best effort */`),
   leaving no diagnostic trail when the cleanup itself fails (EROFS,
   immutable flag, sandboxed container). Added a `createDebugLogger`
   import (`'ATOMIC_WRITE'` category) and a debug log so incident
   response can correlate the original write error with a subsequent
   EEXIST loop. Async + sync.

3. The pre-open `unlink(targetPath)` correctly propagates non-ENOENT
   errors (EACCES on parent dir, EROFS), but no test exercised that
   path. Added an `unlink` field to the `_testFs` seam (async + sync,
   matching the existing `rename` / `writeFile` / `chmod` / `fchmod`
   pattern) and two tests verifying EACCES propagates instead of
   getting hidden behind a downstream EEXIST from O_EXCL.

4. `resolveSymlinkChain(filePath)` ran before the function's main
   try-block, so symlink-resolution errors (EACCES on intermediate
   dir, ELOOP from circular chain) bypassed the
   `atomicWriteFile("path"): ...` annotation that every other failure
   path applies — leaving `err.path` referencing an internal
   intermediate directory the caller never asked about. Wrapped with
   `.catch(err => throw annotateWriteError(err, filePath))` (async)
   and the equivalent try/catch for sync. Added a real-fs ELOOP
   regression test for both variants (skipped on Windows).

64/64 atomicFileWrite tests pass (was 58).

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core): widen atomicWriteJSON options + tighten trustedHooks perms + cover backoff/mkdir branches

Five PR QwenLM#4333 round-11 review items, all small and bounded:

1. **atomicWriteJSON option-type widening**: Was typed as the narrower
   `AtomicWriteOptions` (retries / delayMs only), so credential-grade
   options added in this PR (`mode`, `forceMode`, `noFollow`) and
   pre-existing ones (`flush`, `encoding`) were silently dropped at the
   type level even though the body spread them at runtime. A future
   maintainer calling `atomicWriteJSON(credPath, creds, {noFollow:
   true, mode: 0o600, forceMode: true})` would have typechecked but
   silently lost noFollow + forceMode + mode. Widened to
   `AtomicWriteFileOptions`.

2. **trustedHooks 0o600 + forceMode**: This file lists user-approved
   *executable hook commands* — strictly more sensitive than the
   sibling state files (`trustedFolders.json`, `tipHistory.json`) that
   already use `{mode: 0o600, forceMode: true}`. Was dropping to the
   process umask (0o644 by default), and a backup-restored looser mode
   was never healed. Now matches the sibling pattern.

3. **renameWithRetry exponential-backoff coverage**: Existing tests
   covered retry count and error propagation but not the
   `delayMs * 2 ** attempt` curve itself. A regression to linear,
   constant, or — worst — regressive backoff (which intensifies under
   Windows AV-scan stress) would have passed every existing test.
   Added a test using `vi.useFakeTimers()` that records gaps between
   mock-rename invocations and asserts `[delayMs, 2*delayMs,
   4*delayMs]` for `(retries=3, delayMs=50)`.

4. **jsonl-utils write() parent-dir creation**: The other `write()`
   test targets a path inside the pre-created `tmpRoot`, so the
   `!existsSync(dir) → mkdirSync(dir, {recursive: true})` branch was
   never exercised. Added a one-liner test that targets a deeply
   nested non-existent path.

5. **writeLineSync docstring accuracy**: The docstring claimed "uses
   a simple flag-based locking mechanism (less robust than async
   version)" but there is no flag-based locking — and `writeLine`
   serializes via per-file `Mutex` that this function bypasses. Now
   accurately documents the lack of locking, the bypass, and the
   `flush: true` rationale (closes QwenLM#3681).

Test results: 65/65 atomicFileWrite tests pass (was 64), 26/26
jsonl-utils tests pass (was 25). Typecheck clean.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core): force JS slow path for appendFile flush + correct debugLogger format

Two PR QwenLM#4333 round-12 review items, both real bugs:

1. **flush:true was silently a no-op for string payloads on appendFile**.
   Node's C++ fast path (binding.writeFileUtf8) for string + utf8 +
   appendFile bypasses the JS-side flush/fsync logic entirely —
   empirically string+flush:true takes ~0.05ms/op (identical to no
   flush) while Buffer+flush:true takes ~4.9ms/op (91× slower,
   proving fsync only runs for Buffer payloads). The data still
   reaches the kernel page cache (the syscall is synchronous), so
   `kill -9` is fine, but power-loss durability — the actual QwenLM#3681
   guarantee — was silently absent.

   Fix: pass `Buffer.from(line, 'utf8')` to both writeLine (async)
   and writeLineSync. This forces the JS slow path that honors
   `flush: true` and actually fsyncs the file. Updated the JSDoc
   on both functions to document the C++ fast-path bypass so a
   future maintainer doesn't revert to the simpler string form.

2. **`debugLogger.debug` doesn't do printf substitution**.
   `debugLogger`'s `formatArgs` (debugLogger.ts:67-77) just joins
   args with spaces — no `util.format()`. The round-10 calls used
   `'orphan unlink failed for %s: %s'` which rendered the literal
   `%s` markers in the log:
       orphan unlink failed for %s: %s /path/to/target Error: EACCES
   instead of:
       orphan unlink failed for /path/to/target: Error: EACCES
   Switched both async and sync sites to template literals, matching
   every other `debugLogger` call site in the codebase.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core): narrow tryChmod catch + writeFileSync(fd) + cover credential write failures

Three PR QwenLM#4333 round-13 review items + one comment-wording softening:

1. **`tryChmod`/`tryChmodSync` catch narrowed (wenshao)** — was bare
   `catch {}` swallowing all errors, while the round-8 `fchmod` catch
   was already narrowed to ENOSYS/ENOTSUP. Same security rationale
   applies — and *specifically* on the EXDEV non-noFollow fallback,
   `tryChmod(targetPath)` is the *sole* mode-setting mechanism for an
   existing target (writeFile ignores `mode` when the target exists),
   so a silent EPERM/EROFS would leave credentials at the old mode.
   Non-credential callers don't pass `mode` → `desiredMode === undefined`
   short-circuits, so they're unaffected.

2. **Sync EXDEV `writeSync` → `writeFileSync(fd)` (yiliang114)** —
   `fsSync.writeSync(fd, buf)` returns bytes-actually-written and can
   short-write; the current code ignored the return, so a partial write
   would silently truncate the credential file with the call still
   returning success after fsync+fchmod. Switched to
   `fsSync.writeFileSync(fd, buf)` which loops internally per Node spec.
   The async sibling (`fd.writeFile`) already handles short-writes;
   this brings sync parity.

3. **`file-token-storage` failure-path coverage (wenshao)** — both
   `setCredentials` and `deleteCredentials` propagate `atomicWriteFile`
   rejections (no try/catch around the call), but no test exercised
   that path. Added two tests mirroring the round-1 sharedTokenManager
   precedent: ENOSPC on `setCredentials` and EROFS on `deleteCredentials`
   both rethrow.

4. **Round-12 comment wording softened (wenshao verification report)** —
   strace on Node v22/v24 confirms string + utf8 + flush:true does
   fsync correctly today, counter to my round-12 "silent no-op"
   framing. Buffer is still the safer documented form (forward-compat
   insurance against any future fast-path optimization), but the
   commit's claim that it was *fixing* a confirmed bug overstated what
   reproduces. Reframed both writeLine and writeLineSync comments
   accordingly without changing the code behavior.

Test results: 109/109 affected suites pass (atomicFileWrite 65,
jsonl-utils 26, file-token-storage 18). Broader credential/state
suites also green: 216/216 across sharedTokenManager + oauth-token-storage
+ qwenOAuth2 + logger + trustedFolders. Typecheck clean.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core): route orphan unlink through seam + cover tryChmod + trustedHooks tests

Three PR QwenLM#4333 round-14 review items:

1. **Path-level tryChmod / tryChmodSync catch narrowing had zero direct
   coverage**. Round-13 narrowed the bare `catch {}` to ENOSYS/ENOTSUP
   only (matching the round-8 fd-level fchmod narrowing), but every
   existing EXDEV test passed `options: undefined`, so `desiredMode ===
   undefined` short-circuited before any chmod was attempted. A
   regression that inverted the catch condition (missing `!` prefix)
   would silently swallow EPERM on every sync credential write
   (trustedHooks, tipHistory, trustedFolders, etc.) with zero
   diagnostic signal. Added 6 tests via the existing `_testFs.chmod`
   seam: parameterized ENOSYS/ENOTSUP swallow + EPERM propagation, for
   both async (`atomicWriteFile`) and sync (`atomicWriteFileSync`).

2. **Orphan-cleanup unlink on the noFollow EXDEV failure path was
   using raw `fs.unlink` / `fsSync.unlinkSync` instead of the
   injected `unlinkImpl` seam**. The pre-open unlink correctly used
   the seam, but the round-9 orphan cleanup added later bypassed it,
   making it the only fs operation in `atomicWriteFile` not flowing
   through the test seam. Routed both async and sync orphan cleanup
   through `unlinkImpl`, and added 2 tests that inject a spy and
   assert orphan cleanup is invoked against targetPath after a
   simulated fchmod EPERM.

3. **`trustedHooks.ts` had no test coverage**. Round-11 migrated it to
   `atomicWriteFileSync` with `{ mode: 0o600, forceMode: true }` —
   strictly the most security-sensitive write in the PR since the
   file stores user-approved executable hook commands — but unlike
   the sibling files (trustedFolders, tipHistory) it had no test
   file. A regression that dropped `forceMode: true` or weakened
   the mode would have passed all existing tests. Created
   `trustedHooks.test.ts` covering: write goes through
   atomicWriteFileSync with `{ mode: 0o600, forceMode: true }`,
   write targets the global qwen dir path, the persisted content
   matches the hook key derived from the hook config, and round-trip
   trust/untrust behavior.

Test results: 73/73 atomicFileWrite tests (was 65) + 5/5
trustedHooks (new). Typecheck clean.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core): add noFollow to trustedHooks/trustedFolders + route tmp cleanup through seam

Two PR QwenLM#4333 round-15 review items:

1. **`trustedHooks.ts` and `trustedFolders.ts` were missing `noFollow:
   true`**. The credential write sites
   (`sharedTokenManager`/`oauth-token-storage`/`file-token-storage`)
   all pass `{ mode: 0o600, forceMode: true, noFollow: true }` to
   prevent pre-placed symlink attacks. The trustedHooks comment
   already called it "strictly more sensitive than trustedFolders /
   tipHistory" — yet credential paths got symlink protection it
   didn't. A pre-placed symlink at `~/.qwen/trusted_hooks.json` (or
   `trustedFolders.json`) could redirect the atomic write to an
   attacker-controlled path, either leaking the executable-trust
   list / trusted-folder list, or leaving the user's real config
   silently stale. Added `noFollow: true` to both write sites and
   updated the assertions in `trustedHooks.test.ts` and
   `trustedFolders.test.ts`.

2. **Tmp-file cleanup at L240 (async) and L568 (sync) used raw
   `fs.unlink` / `fsSync.unlinkSync` instead of the injected
   `unlinkImpl` seam**. Pre-open unlink and orphan cleanup correctly
   routed through `unlinkImpl`, making the tmp-cleanup branch the
   only outlier — `_testFs.unlink`-injecting tests couldn't intercept
   this path, weakening the seam abstraction. Behavioral impact is
   nil (cleanup is best-effort, errors swallowed), but consistency
   matters for future test authors. Routed both async and sync
   variants through `unlinkImpl`.

Test results: 99/99 affected suites pass (atomicFileWrite 73,
trustedHooks 5, trustedFolders 21). Typecheck clean on core.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core): use path.join in trustedHooks test to fix Windows CI

PR QwenLM#4333 round-15 review item (Critical, from claude-opus-4-8 /qreview):

trustedHooks.test.ts:68 hardcoded a POSIX path
('/mock/home/.qwen/trusted_hooks.json'), but production builds the path
with path.join(Storage.getGlobalQwenDir(), 'trusted_hooks.json')
(trustedHooks.ts:29-31). On Windows path.join emits '\' separators, so
the mocked atomicWriteFileSync receives
'\mock\home\.qwen\trusted_hooks.json' and the toBe assertion fails — the
cause of the red Test (windows-latest, Node 22.x) check. macOS/Linux
runs are green because forward slashes match.

Build the expected path with path.join so it matches the platform
separator, and add the node:path import.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* fix(core,cli): round-16 review — tipHistory noFollow + atomicWrite test coverage + LSP concurrency note

Four PR QwenLM#4333 follow-up review items from wenshao:

1. tipHistory.ts was missing noFollow: true while the sibling
   0o600+forceMode config sites (trustedFolders.ts:192,
   trustedHooks.ts:63) both set it. The cleanup commit added forceMode
   "for consistency with other 0o600 sites" but stopped short of
   noFollow. All three store user-trusted paths and should share the
   same pre-placed-symlink protection. Added noFollow: true.

2. annotateWriteError idempotency guard had no direct coverage. The
   guard switched from includes(targetPath) to startsWith(fnName+"(")
   because real syscall errors embed the *tmp* path (which contains the
   target as a substring), so the old guard silently skipped annotation
   on every real failure. Added a test where the rename error message
   embeds a tmp-style path containing the target; the correct startsWith
   guard still annotates it, and reverting to includes would fail it.

3. non-EXDEV rename failure tests only asserted /EIO/, not the
   atomicWriteFile(...): / atomicWriteFileSync(...): annotation prefix
   that the production re-throw applies on that path. Tightened both
   async and sync assertions to match the prefix.

4. applyTextEdits concurrency constraint was undocumented. The async
   read-modify-write has await points between read and write;
   atomicWriteFile prevents torn files but does not serialize writers,
   so concurrent same-path edits can lose updates. Latent today (no
   production caller / no workspace/applyEdit handler), so documented
   the per-file-serialization requirement instead of adding a lock.

Not taking: extending the _testFs seam with open/openSync to fault-inject
O_EXCL EEXIST — the noFollow/O_EXCL security behavior is already covered
by the real-fs behavioral tests (noFollow EXDEV fallback refuses to
follow symlinks); a seam-injected EEXIST would only exercise error
propagation, not the guarantee.

Test results: atomicFileWrite 74/74 pass, eslint clean.

Refs: QwenLM#4333, QwenLM#4095 Phase 2

* test(core): route atomicWriteFile open through _testFs seam + assert O_EXCL (PR QwenLM#4333 review)

🤖 Generated with [Qwen Code](https://github.com/QwenLM/qwen-code)

…M#4699)

* docs: add /diff command documentation to commands.md

Add section 1.8 documenting the /diff interactive diff viewer,
including source picker (Current + per-turn diffs), keyboard
shortcuts, dialog example, and non-interactive mode output format.
Also add /diff entry to the 1.2 Interface and Workspace Control table.

* docs: add auto theme detection section to themes.md

Document the 'auto' theme setting and its detection fallback chain
(COLORFGBG → OSC 11 → macOS system appearance → default dark),
including notes for tmux/SSH environments.

* docs: fix checkpointing default description in /diff section

Checkpointing defaults to false, not true. Updated from
"on by default" to "disabled by default" per reviewer feedback.

* docs: fix file checkpointing default in /diff section

File checkpointing (used by per-turn diffs and /rewind) defaults to
enabled in interactive mode. Session checkpointing (/restore) is the
one that defaults to disabled. Corrected the description accordingly.

---------

Co-authored-by: 克竟 <[email protected]>

* fix(cli): unblock third-party model setup for MiniMax-M3

Keep CLI provider setup focused on issue QwenLM#4663 by combining free-form model entry with searchable recommended selections, preserving custom-provider multi-model entry, and carrying saved credentials through model switches.

Constraint: Issue QwenLM#4663 requires MiniMax-M3 metadata, searchable recommended model selection, and manual model IDs without VS Code scope changes.

Rejected: VS Code auth-flow edits | User narrowed scope to CLI-only behavior.

Confidence: high

Scope-risk: moderate

Directive: Keep Custom Provider multi-model entry separate from built-in provider recommended-model selectors.

Tested: cd packages/cli && npx vitest run src/config/auth.test.ts src/ui/auth/ProviderSetupSteps.test.tsx src/ui/components/ModelDialog.test.tsx src/ui/components/shared/TextInput.test.tsx

Tested: cd packages/core && npx vitest run src/core/modalityDefaults.test.ts src/core/tokenLimits.test.ts src/models/modelRegistry.test.ts src/models/modelsConfig.test.ts src/providers/__tests__/presets/minimax.test.ts src/providers/__tests__/provider-config.test.ts

Tested: git diff --check; npm run lint; npm run typecheck; npm run build

Not-tested: Full package test suite on Windows due existing symlink permission / unrelated failures noted in review.

* test(cli): prevent Windows CI races in prompt suggestion submit

Constraint: Windows CI can lag Ink/React render settling after follow-up debounce.

Rejected: Longer real-time sleeps | still flaky and slower under runner load

Confidence: high

Scope-risk: narrow

Directive: Prefer timer-driven state transitions over fixed sleeps in InputPrompt tests.

Tested: cd packages/cli && npx vitest run src/ui/components/InputPrompt.test.tsx

Not-tested: Full cross-platform CI matrix

* fix(core): Repair duplicate tool call IDs

Co-authored-by: Qwen-Coder <[email protected]>

* fix(core): Preserve streaming tool replay metadata

Avoid mutating completed streaming tool calls when a provider replays the same tool call ID with another arguments chunk. This keeps the first surviving call metadata and buffer intact while preserving fragmented JSON accumulation before completion.

Co-authored-by: Qwen-Coder <[email protected]>

* fix(core): Address duplicate call id review feedback

Use the shared tool-call dedupe helper from ACP Session, preserve distinct empty call IDs in scheduler and non-interactive batches, and log dropped duplicate scheduler requests for diagnosis.

Co-authored-by: Qwen-Coder <[email protected]>

* fix(core): Address duplicate tool call review feedback

Co-authored-by: Qwen-Coder <[email protected]>

---------

Co-authored-by: Qwen-Coder <[email protected]>

…#5110)

The /copy command supports code block selection with language filtering,
LaTeX, and Mermaid — but the TUI autocomplete only showed "[N]" with a
description about copying the full AI reply. Users had no way to discover
these capabilities from the prompt.

Update argumentHint to "[N] [<lang>|code|latex|mermaid] [<index>]"
and description to mention all supported targets, noting N counts from
the last message.

…LM#5109)

* feat(web-shell): collapsible TodoWrite history with status diff

Inline todo_write updates rendered as a generic, non-collapsible tool row
crammed in with surrounding tool calls — the todo-specific renderer was
effectively dead code because the detector matched the literal "todowrite"
while the wire name is "todo_write" (kind "think").

- Detect the todo tool by name (todo_write / todowrite) instead of relying
  on the unrelated tool kind, reviving the rich rendering and also
  populating the floating todo panel that was empty on the daemon path.
- Render each update as its own standalone group, collapsed by default to
  the per-snapshot diff (just-completed / just-started items) or the current
  step, expanding to the full checklist; the header shows completed/total.
- Use consistent status glyphs (●/◐/○) in both collapsed and expanded views
  via a shared TodoView component used by ToolGroup and PlanMessage.

* fix(web-shell): scope todo diff per task identity; address review

- [Critical] computeTodoTimeline keyed its running state on todo.id alone, but
  ids aren't globally unique (ACP assigns positional ids, models renumber per
  plan), so a later plan diffed against a previous plan's stale terminal status
  and silently dropped events in the collapsed view. Key on id+content so
  distinct tasks stay separate; add an id-reuse regression test.
- Stabilize the TodoTimelineContext value with a signature-cached Map so
  streaming ticks that don't touch a todo snapshot no longer re-render every
  todo/plan row.
- Drop the unused completed/total from TodoSnapshotDiff (consumers compute the
  count locally — single source of truth).
- Fall back to the raw result summary when a todo_write payload is unparseable.
- Add PlanMessage rendering tests and extractTodosFromToolCall coverage; remove
  stale "step time" comments left after dropping per-step timing.

* test(web-shell): document todo-diff keying limits; cover signature

Follow-up to review on the id+content keying in computeTodoTimeline:
- Document the two rare trade-offs in the todoStateKey doc (a mid-task reword on
  a stable id, and unrelated plans reusing both id and content), both degrading
  to "the collapsed diff omits one event" while the expanded list stays correct.
- Pin behavior with tests: an item carried over and completed in a later turn
  (which id+content handles but a user-turn reset would drop), plus the two
  documented gaps.
- Add todoTimelineSignature tests: stable across non-todo edits; changes on any
  id/status/content change.

* refactor(web-shell): isolate plan context read; expand todo test coverage

Address follow-up review:
- Extract PlanEventSummary as the sole TodoTimelineContext consumer, mirroring
  ToolGroup's TodoToolBody so the memo-shielded PlanMessage stays stable when
  the timeline Map reference changes.
- Note the spurious-`started` axis of the reword trade-off in the todoStateKey
  doc (a reword can drop a completion or emit a stray start).
- Add direct isTodoWriteToolName tests (incl. the `todowrite` ACP variant) and
  a todoTimelineSignature empty-transcript test.

Signed-off-by: Yufeng He <[email protected]>

* fix(core): bound hard rescue compression retries

* fix(core): count rejected hard rescue noops

* fix(core): clarify hard rescue retry accounting

* test(core): cover hard rescue reactive fallback

* test(core): remove duplicate debug logger mock

* fix(core): include hard rescue count in stop log

* fix(core): compress when usage metadata is missing

* fix(core): harden missing compression usage fallback

* test(core): cover missing-usage inflated compression fallback

* fix(core): clarify missing usage compression fallback

* test(core): align compression fallback assertion with summary trailer

* test(core): use ascii cjk fixture escape

* test(core): tighten missing-usage compression assertions

* docs: Refresh daemon developer docs

Co-authored-by: Qwen-Coder <[email protected]>

* docs: Address daemon review feedback

Co-authored-by: Qwen-Coder <[email protected]>

* docs: Address daemon review suggestions

Co-authored-by: Qwen-Coder <[email protected]>

* codex: address PR review feedback (QwenLM#4412)

Co-authored-by: Qwen-Coder <[email protected]>

* codex: address PR review feedback (QwenLM#4412)

Co-authored-by: Qwen-Coder <[email protected]>

* codex: address PR review feedback (QwenLM#4412)

Co-authored-by: Qwen-Coder <[email protected]>

---------

Co-authored-by: Qwen-Coder <[email protected]>

* fix(core): include response tokens in prompt estimate

* fix(core): avoid double-counting reasoning output tokens

* fix(core): restore response token estimate on resume

* test(core): cover output token reset paths

* refactor(core): share output token estimate helper

* refactor(core): seed resume output tokens atomically

* fix(core): tighten resume output token accounting

* fix(core): count equal reasoning tokens conservatively

---------

Co-authored-by: Cursor Agent <[email protected]>
Co-authored-by: JerryLee <[email protected]>

…nected (QwenLM#4894)

* fix(dual-output): prevent FIFO blocking on startup when no reader connected

DualOutputBridge's ENXIO fallback used a blocking createWriteStream on
FIFOs, causing the TUI to hang indefinitely when launched with
`--json-file <fifo>` before a reader connects (issue QwenLM#4727).

Fix: use O_RDWR | O_NONBLOCK for the FIFO fallback path. This POSIX
trick satisfies the kernel's "at least one reader" requirement without
blocking. A buffer high-water-mark (1 MB) self-disables the bridge if
no consumer ever drains the pipe.

Also updates Quick start docs to recommend regular files as the default,
with FIFOs documented as an advanced option that now works without
ordering constraints.

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

* fix(dual-output): address review feedback — guardActive, stream.destroy, tests

- Rename isBufferOverflowing() to guardActive() (command-query separation)
- Apply buffer guard to all write methods, not just processEvent
- Call stream.destroy() on overflow so FIFO consumers get EOF
- Handle destroyed stream in shutdown() to prevent hanging
- Add test: bridge disables on buffer overflow + stream is destroyed
- Fix doc: --input-file requires regular file (not FIFO), stat.size=0

Generated with AI

Co-authored-by: Qwen-Coder <[email protected]>

---------

Co-authored-by: 秦奇 <[email protected]>
Co-authored-by: Qwen-Coder <[email protected]>

* feat(workflow): add Qwen Scheduled Issue Autofix workflow for automated issue handling

* fix(ci): harden scheduled autofix workflow

* fix(ci): harden scheduled autofix recovery

* fix(ci): keep autofix cleanup best effort

…op (QwenLM#5128)

QwenLM#5036 carved the deterministic identical-tool-call check out of the
`model.skipLoopDetection` gate, turning it into a hard-stop that fires
even when loop detection is disabled. Because `skipLoopDetection`
defaults to true (settingsSchema: "to avoid false-positive
interruptions"), this silently re-enabled loop halts for the default
configuration and broke the documented escape hatch — the
non-interactive guidance in nonInteractiveCli.ts told users to set
`model.skipLoopDetection: true`, which no longer disabled the halt and
is unreachable in non-interactive mode (no disable dialog).

Gate both the deterministic and heuristic detector paths behind the
single flag again. The deterministic split, retry-reset, and pending
tool-call splice introduced by QwenLM#5036 still apply once detection is
explicitly enabled (skipLoopDetection: false), so the runaway guard
remains available as opt-in without overriding the default-off contract.

…QwenLM#5138)

…env) (QwenLM#5122)

* feat(computer-use): configurable screenshot max dimension (setting + env)

Add a user-level knob for cua-driver's screenshot longest-edge cap. The
old open-computer-use backend exposed this via OPEN_COMPUTER_USE_IMAGE_*
env vars; the cua-driver migration dropped them, leaving only the
model-driven set_config tool. This restores deterministic user control.

- Setting tools.computerUse.maxImageDimension (number; default -1 = keep
  cua-driver's built-in default of 1568; 0 disables resizing / full
  resolution; a positive value caps the longest edge).
- Env override QWEN_COMPUTER_USE_MAX_IMAGE_DIMENSION (takes precedence
  over the setting; invalid/negative values fall through).
- Resolution lives in resolveMaxImageDimension(); applied via the
  cua-driver set_config tool once per (re)connect in
  ComputerUseClient.doStart — best-effort, never aborts startup, and
  re-applied after a daemon-restart reconnect.
- Docs: document tools.computerUse.{enabled,maxImageDimension} in
  settings.md (the block was previously undocumented). Refresh stale
  ocu/npx comments left in client.ts + install-state.ts by the migration.

Precedence: env var > setting > cua-driver default.

* chore(computer-use): finish ocu→cua-driver cleanup in schema-sync script

The cua-driver migration (QwenLM#5051) left scripts/sync-computer-use-schemas.ts
pointing at the old open-computer-use backend: it npx'd
@qwen-code/open-computer-use, hard-coded the 9-tool ocu surface, and emitted an
"open-computer-use" header. Re-running it — which constants.ts' version-bump
procedure tells maintainers to do — would have clobbered the migrated 35-tool
cua-driver schemas.ts.

- Drive the locally-pinned `cua-driver mcp` binary (binaryPath /
  CUA_DRIVER_VERSION from constants.ts) instead of npx'ing ocu; expect 35
  tools and warn (don't fail) on drift.
- Emit the cua-driver-flavored schemas.ts header.
- Refresh install-state.test.ts fixtures from ocu package specs to the
  cua-driver-rs approval-key form the field actually stores now.

Verified the fixed script reproduces the committed 35-tool surface exactly
(modulo prettier formatting). No dead env-var handling remained — the module
reads only QWEN_COMPUTER_USE_{AUTO_APPROVE,DOWNLOAD_HOST,MAX_IMAGE_DIMENSION}.

* fix(core): bound active tool result history

Co-authored-by: Qwen-Coder <[email protected]>

* fix(core): keep tool result budget defaults lightweight

Move the new tool-result history budget default into a lightweight config defaults module so microcompaction does not load the full Config graph during service tests. Update the ACP worktree test mock to include the public default export used by settings schema imports.

Co-authored-by: Qwen-Coder <[email protected]>

* fix(core): address tool result budget review

Handle negative legacy idle thresholds consistently, clarify size compaction diagnostics for pending tool results, promote successful microcompaction logs to info, and strengthen tests/docs around skipped results and soft thresholds.

Co-authored-by: Qwen-Coder <[email protected]>

* fix(core): log protected tool result overages

Co-authored-by: Qwen-Coder <[email protected]>

---------

Co-authored-by: Qwen-Coder <[email protected]>

…nLM#5118)

* feat(web-shell): per-task token & time detail on completed todos

Expanding a completed task in the todo list now reveals when it ran (start /
end / duration) and what it spent: input / output / cached tokens, API time,
and tool time.

The agent stamps a cumulative-usage snapshot onto each todo (plan) update via
`_meta.stats`; the SDK normalizer carries it into the TodoWrite tool call's
rawOutput, and the web-shell diffs consecutive snapshots for tokens and API
time while summing transcript tool durations for tool time.

Works live (no polling race) and on /resume: tokens and tool time are
reconstructed from persisted usage metadata, while API time is live-only since
per-turn durations are not replayed. Sessions whose agent never stamped a
snapshot degrade gracefully to start/end + tool time.

* refactor(web-shell): show task duration inline on the end-time row

Trail the elapsed duration after the end time as a dimmed parenthetical
("12:34:15 (4m 14s)") instead of a separate row, since it's derived from the
start/end pair.

* fix(web-shell): correct per-task detail for reused todo ids; review follow-ups

- computeTodoDetails: when a completed id+content key restarts as in_progress (positional plan-N ids repeat across plans), reset the window so the new task diffs its own start instead of the prior task's far-earlier boundary — which rendered a cross-plan window with wildly inflated token/time numbers. Correct the todoStateKey JSDoc accordingly.
- Tool time: sort spans once and binary-search the task window instead of an O(todos x spans) scan per completed task.
- Tests: add the reuse-reset case, a windowed tool-time case, an SDK-normalizer -> extractTodoStats contract test (locks the stats passthrough so a field rename fails loudly), and a stopPropagation test (expander click must not bubble to the tool-row header).

* fix(web-shell): reset todo detail window on reopen via pending, not just direct

Track keys that have ever reached 'completed' instead of checking prev === 'completed': a reopened task can pass through 'pending' (completed → pending → in_progress), where prev at the re-activation is 'pending' and the direct check missed it, leaving a stale baseline that diffs across both runs. A pause/resume that never completed (in_progress → pending → in_progress) still keeps its first baseline, so its diff captures the whole task.

* fix(web-shell,cli): harden todo stats against NaN poisoning and partial snapshots

- MessageEmitter: only fold finite usage/duration values into the cumulative accumulator. A NaN/Infinity (incl. a NaN that survives `?? 0`) would poison the running total forever, making every later snapshot fail extractTodoStats and silently show 'not captured' for the rest of the session.
- extractTodoStats: require the token fields but default the live-only apiTimeMs to 0 when absent/non-finite, so a snapshot that omits it keeps its valid token counts instead of being dropped whole.
- computeTodoDetails: gate the start baseline on the stored value, not Map.has — a stats-less start (e.g. a plain plan message) recorded undefined, which Map.has treated as already-set, blocking a later stats-bearing snapshot from upgrading the baseline.
- Document the MessageEmitter-before-PlanEmitter ordering invariant in both emitters.

…ure (QwenLM#5123)

Mermaid diagrams with cylinder shapes [(...)] and CJK/emoji content
failed to render because DOMParser strict XML parsing rejected
XHTML content (e.g. <br>) inside <foreignObject>, causing
sanitizeSvg to return empty string and display 'Mermaid render failed'.

The sanitizeSvg function was redundant — mermaid's securityLevel:'strict'
already uses DOMPurify internally for XSS protection. Removing the
custom sanitizer fixes the rendering issue and simplifies the code.

Co-authored-by: ytahdn <[email protected]>

When enableFollowupSuggestions is true, display the generated
follow-up suggestion as the input placeholder text (replacing
the default "Type your message..."). Tab/Enter/Right arrow
accepts the suggestion; typing dismisses it.

Also change the default of enableFollowupSuggestions from false
to true so the feature is on by default.

Key changes:
- AppContainer: dismissPromptSuggestion no longer clears
  promptSuggestion state, preserving it for placeholder restore
  after user types then deletes
- InputPrompt: Tab/Enter/Right arrow/typing handlers check
  promptSuggestion prop as fallback when followup.state is not
  visible (e.g. after 300ms delay or user dismissed)
- Composer: placeholder shows suggestion text when available
- hasTabConsumer: include promptSuggestion to prevent Windows
  bare Tab from cycling approval mode