2479 Commits

This Branch

This Branch

All Branches

Author	SHA1	Message	Date
DevvynandClaude Sonnet 5	b45a96358e	feat(custom-model): warn before launching Claude on a model too small for its own overhead ... Claude Code's own fixed per-turn overhead (system prompt + tool schemas, ~36.4K tokens measured live) can exceed a small local model's entire real context before any conversation history exists to compact — confirmed live twice as an in:0 out:0 failure on the very first message sent. CLAUDE_CODE_MAX_CONTEXT_TOKENS cannot fix this: it only governs when history gets compacted, and there is none on message one. - exceedsSafeContextFloor() (custom-model-routes.ts): true when a CLI's registry entry declares contextLengthVar (currently only claude) and the model's discovered context is below CLAUDE_MIN_SAFE_CONTEXT_TOKENS (40000). A no-op for every other CLI by construction. - Both apply routes (POST /api/sessions/:id/custom-model and the quick-start customModel path) check this before the swap-conflict check and before launching/restarting anything, returning {requiresContextWarning, modelId, contextLength, minSafeContextTokens} — skipped when confirmed:true. - Frontend: #customModelContextWarningModal + _confirmContextWarning/ _resolveContextWarningConfirm (session-ui.js), wired into both _quickStartWithCustomModelConfirm and _runCustomModelEntryViaRestart (the path Claude actually uses) ahead of the swap-confirmation check. Explains the fix in-modal: give the model an explicit larger -c/ --ctx-size in llama-swap instead of relying on --fit-ctx, which optimizes for the biggest model that fits rather than the biggest context. Tests added for the route-level warning/confirm/skip cases and the frontend modal + launch-flow wiring. Docs updated (custom-model- endpoints.md, wiki/Custom-Model-Endpoints.md) and the PR's running changeset extended. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-17 07:36:57 +08:00
Randalix	29984c639d	fix(remote): stop the flush losing a chunk, and reset the host form's wake fields ... Own review pass over the PR: - `_flush` took the chunk out of the buffer only AFTER awaiting the write. Input arriving during that await is enqueued (`waking` is still set, so it takes the buffer path), and the 4 KB cap then drops the OLDEST chunk — which is the one already on its way to the pane. The `shift()` that followed removed the NEXT chunk instead, so the drop-oldest bookkeeping silently lost a chunk that was never written, while the log line blamed the one that was. The chunk is now removed before the await and re-inserted at the FRONT on a failed write, so the order of the queue behind it is preserved. Regression test: a chunk enqueued during the first write of a full buffer must still reach the pane (red against the old order). - `showCreateCaseModal()` reset the remote-host form fields but not the two new wake inputs, so one host's MAC/command carried over into the next host that form saved. - The banner's pre-poll `wakeConfigured` labelled a command-only host as 'mac'. Nothing reads the distinction, but the field is documented as which path is configured, so it says the truth until the first poll corrects it. - Stale `resolveRemote` comment ("only for sessions that have no usable target of their own"): after the host config became authoritative in both directions it is consulted on the TTL regardless.	2026-09-16 21:06:25 +02:00
Randalix	acb8d4b0aa	docs(remote): correct what the wake PR moved ... - `host-wake-ui.js` joins the documented load order (12.2) and gets its `@dependency`/`@loadorder` tags; the frontend module count is 33, not 32. - `remote-wake` is not "(pure)" — the module uses `dgram`/`net`/`child_process`. - SSE counts: 160 constants, and the category is "Remote auto-reconnect / wake (5)"; the route table's per-file counts are refreshed (sessions 37, cases 34). - The CLAUDE.md wake rule now names the create/attach wake, the 40 s request budget, the whole-chunk paste drop, the registry's lifetime (drop on cleanup, stop on shutdown) and the deliberately non-wake-aware WebSocket keystroke path — that paragraph is what the next person reads. - Reverted the eight lines of unrelated Prettier markdown churn in `docs/architecture-invariants.md` (docs/ is not in the format glob, so it was an editor): only the new wake paragraph remains in the diff.	2026-09-16 20:44:48 +02:00
Randalix	7b947fa3f1	fix(remote): close the wake-state leaks and the dishonest wake budget ... Review follow-up on the wake-on-LAN PR (five findings, all of them about the state the feature keeps and the budgets it inherits): - Wake state is dropped by `WebServer.cleanupSession` instead of the two delete routes, so it now goes with the session on EVERY cleanup path (cron, admin, scheduled-run teardown, error paths) instead of surviving with up to 4 KB of the user's buffered keystrokes. `registerSessionRoutes` returns the registry so the server can own its lifetime without the wake-capable code living in `server.ts`; the wiring guard is updated to allow that and gains a second assertion that `server.ts` calls nothing but `drop`/`stop` on it. - `_effectiveRemote` returns before `_state`, so a LOCAL session no longer gets a wake-state entry — the input gate runs on every keystroke, so that entry used to be allocated for every session the user types in. - An input chunk larger than the 4 KB cap is dropped OUTRIGHT instead of being head-trimmed and then written as a fragment: one paste is one `input` value and was never typed character by character, so its tail is a partial command the user never sent. The drop is logged. - The manual wake button passes `REMOTE_WAKE_REQUEST_READY_TIMEOUT_MS` (40 s) like the create/attach paths, instead of inheriting the 90 s session default that the dashboard's reverse proxy cuts off at 60 s. - `RemoteWakeRegistry.stop()` aborts in-flight readiness polls (abortable sleep) and refuses new wakes, and `WebServer.stop()` calls it, so a restart during a wake no longer waits the poll out. - The banner/toast wording keys off a new `queuedInput` flag on the two SSE events, which is true only when the server actually holds bytes: browser keystrokes travel over the WebSocket, which never passes through the registry, so the wake BUTTON must not promise queued input. The failed-wake path also stops pattern-matching the error message (it re-asks the reachability route) and the WoL dialog says "admin-only" instead of "host not found" for a non-admin in multi-user mode.	2026-09-16 20:44:39 +02:00
Michael GrundbergandClaude Opus 5	39976041e0	fix(sessions): let a dismiss reach the entries a restore is holding ... Fourth review of the reboot-restore branch, and the third to find a defect in the previous round's fix. This one is the same shape as its predecessor: a counter keyed on one thing, compared against a set keyed on another. The generation counter was indexed by the entry's owner, while the in-flight set holds the caller doing the restoring. Those are the same person exactly when a user restores their own sessions, which is every case the tests covered. The route deliberately supports the other case: an admin may spend another user's entries. So when an admin restored Bob's sessions and Bob dismissed the banner, nothing matched, the entries came back, and a plan Bob had explicitly dismissed was re-armed for another twenty-four hours. Rather than reconcile the two key spaces, the counter is gone. `take()` now parks the entries it hands out, remembering which caller is spending them, and they stay parked until that restore ends. A dismiss filters the parked entries by `canAccess(entry.owner)` — the same predicate it already applies to the plan — so it reaches them wherever they are. `releaseFlight()` puts back only what is still parked. Expiry and a fresh boot plan unpark everything, for the same reason. There is one key space now, the entry's owner, and the spender is only ever used to tell two concurrent flights apart. That removes `generations`, `snapshotGenerations()`, `bump()`, `bumpAll()` and the argument threaded through the route. The discard grew the teardown it still lacked. A rebuild can fail after startInteractive() resolved, and a restored workspace still carries Codeman's hooks, so the CLI can post a hook event within milliseconds; the transcript watcher that starts from it, the attachment registry, the wait registry and the approvals inbox all outlive the listeners and would meet the retry, which reuses the session id by design. Its steps also run in reverse order now, so no live listener can reach a tracker that has already stopped, and the mux kill has its own guard, because stop() kills the pane in its last block after destroying four trackers. Tests. The run-summary test named an interval and asserted a map entry, so dropping stop() left it green; it now spies on stop(). Nothing pinned that before-spawn must precede setupSessionListeners, which reads the flag that phase restores, so swapping the two lines was silent; the ordering test now includes the listener setup. The retry assertion was a tautology and now asserts a different refs object. Both strengthened tests were verified by reverting their fix. Two new tests cover the admin-restores-another-owner cases this round was about. The server in the discard test is built once and stopped, since its constructor registers handlers on module-level watchers, and the workspace is removed through safeRmHomeTree. Refs #411 Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-09-16 16:51:00 +02:00
Michael GrundbergandClaude Opus 5	71ed7b127c	fix(sessions): make the discard a real inverse of the construction ... Third review of the reboot-restore branch. The narrow discard the previous commit introduced avoided everything cleanupSession() did wrongly, and in dropping so much of it also dropped four things it had to keep. The worst broke the retry the whole design rests on. setupSessionListeners() returns early while sessionListenerRefs still holds the session id, and the discard never cleared that entry. So the advertised flow — a rebuild fails because the agent binary is missing, the user fixes their PATH and clicks again — reused the same id, wired no listeners at all, and produced a tab that never showed output, never updated its status and never persisted. That is worse than the leak the discard was added to prevent. Three more registrations leaked with it: a RunSummaryTracker and its interval, an image watcher on the workspace, and the Ralph fix-plan watcher. The discard now undoes each registration setupSessionListeners() makes, in its order, and the per-session custom-model config directory, which holds the endpoint's API key literally and which nothing else would ever remove. The image-watcher flag was restored after the code that reads it, so a session came back reporting the feature as on with nothing watching. It moves to the before-spawn phase, and that phase now runs before the listeners rather than after them. The generation counter that lets a mid-restore dismiss win was global while clear() is ownership-scoped, so one user's dismiss discarded another user's unspent entries, permanently, because nothing rebuilds an in-memory plan. It is now per owner. Bumping only the owners of entries the dismiss removed was not enough either: take() has already emptied the plan by then, so a dismiss landing mid-restore saw nothing of that owner's to remove and invalidated nothing. The owners that matter are those with a restore in flight, filtered by what the dismissing user may access, and that is what clear() now bumps. Plan expiry bumps too, so a restore straddling the 24-hour boundary cannot hand entries back and give an expired plan another full day. Tests. discardPartiallyBuiltSession had no test at all: the only implementation any test ran was the mock's one-line stub, which is why every defect above was invisible. test/discard-partially-built-session.ts drives the real WebServer, and the retry assertion fails if the listener refs are left behind — verified by reverting the fix. The dismiss-race test drove the registry by hand, so deleting the route's generation argument left it green; it now goes through the route, and two further tests cover the multi-user cases. The mock context has now gone stale twice, because route tests pass it as `ctx as never` and tsconfig.json includes only src, so nothing ever compares it to the ports. A type-level guard is therefore inert — I wrote one and confirmed it never fires. test/mocks/mock-route-context-completeness.ts compares the mock's keys against WebServer.createRouteContext() at runtime instead, and names what is missing. Also: the API reference now says workspace-forbidden is judged against the owner's grant, the banner's module header no longer claims Restore always dismisses it, and the detail span gets the same min-width: 0 the phone rule already needed. Refs #411 Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-09-16 15:34:47 +02:00
Michael GrundbergandClaude Opus 5	fa52753e8b	fix(sessions): undo a failed rebuild without deleting the user's data ... A second review of the previous commit found that its own repair for the session leak introduced three defects, all from reaching for cleanupSession() to undo a half-built session. That function is the user-initiated delete, not an undo. It banked the session's historical token and cost totals into the lifetime figures, and a reboot never runs cleanup, so those totals had never been counted before; every failed rebuild added them again. It saw the pin that had just been restored and demoted the record to `stopped`, which this pass reads as the durable marker of a deliberate kill, so a pinned session whose rebuild failed became permanently unrestorable. And it recursively removed `.claude-images` from the working directory, which belongs to the workspace rather than to the session, so a failed rebuild destroyed the pasted images of any other live session in that repo. discardPartiallyBuiltSession() now undoes only what the construction did: the map entry, the tab-layout slot, the listeners and any pane the launch created before throwing. The persisted record, the lifetime totals, the Ralph state and the workspace's files are left alone. Re-applying the persisted state also splits in two, which removes the first two defects at the root rather than only at the call site. The half that shapes the pane, the custom-model environment and the nice priority, still runs before the spawn. The half that is the session's own history now runs after it, so a session whose pane never started carries no totals and no pin for anything downstream to misread. The rest of that review. The multi-user workspace confinement re-check read the requesting user's grant, and returns true for an admin, so the case its own comment described was the one it missed; it now resolves the entry owner's grant through isWorkingDirAllowedForUsername, the way cron does. A forbidden workspace goes back on offer, matching both the registry's stated contract and the API reference. The client re-reads the plan after a restore instead of blanking the banner, so entries the server put back stay reachable, and a 409 now says a restore is already running rather than reporting a failure. A dismiss arriving mid-restore wins, through a generation counter the route carries across its take. The re-application also restores the tab colour, the image-watcher flag and the original pinnedAt, via a new Session.restorePin that does not re-stamp the pin time. The phone breakpoint gains min-width: 0, without which a nowrap flex item never shrinks and the buttons still overflow, and it folds into the existing phone block. Ralph's loop configuration still does not survive a restore, because toState() reads it off a live tracker and there is no way to keep it without arming the loop. The method now says so rather than leaving it implied. Tests. The capacity test could not fail on the property it existed for: it filled the board past the cap before the loop, so a single pre-loop check would have passed it. It now leaves one seat, so only a per-iteration check restores exactly one entry. New tests cover the ordering around the spawn, a throw before the loop returning the whole plan and releasing the flight, the dismiss-during-restore race, and that the failure path calls the narrow discard rather than the delete. The shared mock context gains the port method it was missing, which is what made the first run of these tests fail for the wrong reason. Refs #411 Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-09-16 15:04:35 +02:00
DevvynandClaude Sonnet 5	993710263d	fix(custom-model): stop trusting /props's n_ctx, parse the real context size from /running's cmd ... Root cause of the context-overflow regression reported live: "API Error: 400 request (36437 tokens) exceeds the available context size (16384 tokens)". Discovery had stored modelContextLengths.qwen3.8-27b-ud-q4_k_xl = 154112, so CLAUDE_CODE_MAX_CONTEXT_TOKENS told Claude Code it had a huge window and it never compacted - but the real llama-swap server was launched with --fit-ctx 16384 (confirmed against /running's own cmd field) and refused the request right at that real limit. /props?model=<id>'s n_ctx (the field discovery read) is confirmed live to be unreliable for a --fit-ctx-launched backend: it reported 154112 for the same model /running says was launched with --fit-ctx 16384 - appears to report the model's theoretical/trained maximum context, not the runtime- configured one. discoverModels() now parses the REAL configured size straight out of llama-swap's own launch command instead (parseCtxFromCmd(), reading /running's cmd field - --fit-ctx first, then the plain llama.cpp -c/ --ctx-size a hand-written command might use), and only falls back to the old /props probe when cmd states no recognizable flag at all. One /running call now covers every loaded model's context length in a single request, same as it already did for the swap-conflict check and the load trigger. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 20:47:12 +08:00
DevvynandClaude Sonnet 5	7bbe408e44	feat(custom-model): live countdown on the loading banner; timeout is now an error ... The loading banner now shows a live countdown against its own timeout (updated every poll, so every second by default) instead of a static "this can take a while" — e.g. "Loading qwen3.8-27b (16.4 GB, typically ~1-3 min) on llama-swap - 47s remaining". If the countdown reaches zero and the model still isn't ready, this is now treated as a real failure rather than a "keep waiting" shrug: - The banner turns into a sticky error (_showCenterStatus gains a `type` option - 'error' drops the spinner and adds a close button, since nothing is "in progress" anymore and a sticky message needs a way to dismiss it), naming the llama-swap server's own logs as where to look for detail. - The session that load was for is closed automatically (closeSession) - requested explicitly: a console left open and pointed at a model that never finished loading is worse than no console at all. Both apply paths now thread the new session's id through to _watchLlamaSwapLoading for this (new required 3rd parameter, after endpointId/modelId). _watchLlamaSwapGeneration's existing stale-call guard extends naturally to this: a superseded call's own eventual timeout recognises it no longer owns the banner and neither shows the error nor closes a session that may by then belong to a different, newer launch. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 20:22:02 +08:00
DevvynandClaude Sonnet 5	55dae31530	feat(custom-model): estimate model load time from its discovered size ... Discovery now also parses a GB figure out of an auto-discovered model's own description (llama-swap writes "Auto-discovered 16.35 GB - parameters auto-fitted by llama.cpp"), stored per model as modelSizesGB - unlike context length this needs no /props probe (the figure is right there in /v1/models) so it is populated for every model regardless of loaded state. A hand-configured profile's own description has no such figure and correctly gets no entry. The loading banner (_watchLlamaSwapLoading) now looks this up and, when known, shows it plus a rough estimate from a small size->time matrix (_estimateModelLoad/_MODEL_LOAD_TIME_MATRIX, session-ui.js) - "Loading qwen3.8-27b-ud-q4_k_xl (16.4 GB, typically ~1-3 min) on llama-swap... this can take a while" - and uses that same estimate's own bracket to scale the banner's default give-up timeout for a very large model, instead of a flat 5 minutes for everything. Explicitly labelled as an UNMEASURED, typical-hardware estimate in every relevant comment - this is not benchmarked against any real endpoint's actual storage/GPU, just a reasonable expectation-setter. A model with no discoverable size (a hand-configured profile) gets no size/estimate shown at all, matching the "never a guess" convention modelContextLengths already established. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 19:58:46 +08:00
DevvynandClaude Sonnet 5	0af233c96c	fix(custom-model): poll llama-swap readiness every 1s, check immediately, extend the cap ... Reported: the "Loading..." banner stayed up past 2 minutes even though llama-swap itself had already finished loading the model. Three fixes: 1. pollIntervalMs default 3000ms -> 1000ms (as asked). 2. The loop now checks readiness IMMEDIATELY on entry rather than sleeping a full interval first - a model that's already ready (a fast load, or a re-apply onto one already loaded) shouldn't sit on "Loading..." at all. 3. maxWaitMs default 120000ms (2 min) -> 300000ms (5 min): a large (20GB+) model reading from disk can genuinely take longer than 2 minutes, which would have looked identical to the reported symptom - "still stuck past the point it should have resolved" - except it would have actually flipped to a "still waiting" warning toast at the 2-minute mark rather than staying on "Loading" indefinitely, so this alone doesn't explain what was reported, but is a real, separate improvement worth making. Also fixes a real, separate bug this surfaced while reasoning through the report: _showCenterStatus's banner is ONE shared, reused DOM node. A second call to _watchLlamaSwapLoading (e.g. switching models again before the first switch's loop had finished) would take over that shared banner, but the FIRST loop was still running and would eventually dismiss or overwrite it once ITS OWN deadline or readiness check resolved - clobbering whatever the second, current loop had put there. A generation counter (_watchLlamaSwapGeneration) now lets each call recognise when it no longer owns the banner and stop touching it silently, rather than only the last call to actually start ever safely reading or writing it. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 19:44:37 +08:00
Michael GrundbergandClaude Opus 5	fbede5cd2a	fix(sessions): act on the dual review of the reboot-restore route ... Fifteen findings from two independent reviews of #442, three of them blocking. Every one is addressed here. The three blockers all sat in the restore route. A rebuild that threw after addSession left a registered session with no pane behind it, visible on the board, holding a layout slot and written to state.json, with its plan entry already spent; the catch now cleans the session up and puts the entry back. The loop checked neither the global nor the per-user session cap, so one click could take a board past a documented limit; capacity is now re-checked per iteration, because the loop is itself creating the sessions it counts. Worst of the three, a rebuilt session carried none of the state its constructor has no parameter for and then persisted itself over the record that held it, zeroing token and cost totals and dropping the pin. The pin matters most: pruning keeps a record only while it is pinned, so discarding it handed the record to the next stale sweep. A new reapplyPersistedSessionState() on the session port restores the pin, the token totals, auto-compact, auto-clear, auto-resume, nice priority, the flicker filter and the custom-model selection, and it runs before both startInteractive and the first persist. The rest, in the order they bite a user. Every rebuild failure was reported as workspace-missing, so the banner told users their repo was gone when the agent had simply failed to start; there are now distinct reasons, and the toast names each one. The client read restored and skipped off the outer response object rather than through the uniform envelope, so every count came back zero and neither toast ever fired. A board left open across the reboot never learned an offer existed, because the banner was seeded only on the page-load path; it now re-reads on every SSE init. The workspace check was existence-only, skipping the multi-user confinement that the create route applies, so a withdrawn grant would not be noticed. The banner had no phone breakpoint while its text was nowrap and its buttons could not shrink. Smaller: a missing workspace is now re-offered rather than dropped, while an already-open conversation is dropped rather than re-offered forever; a throw anywhere in the route returns the unspent entries instead of discarding the plan; the single flight is keyed by owner, since take() already stops two callers receiving one entry; the env clamp's header no longer claims a protection it cannot provide on this path today, and names the check that does bite; the three endpoints are documented in docs/api-reference.md; and the module header now says that os.uptime() reads the host's clock, so the feature is effectively off inside a container. The review also explained why the tests missed all of this: they proved the construction claim through their own copy of the construction rather than through the route, and the route tests used workspaces that did not exist, so no Session was ever built. test/routes/reboot-restore-rebuild-failure.ts mocks the Session module to drive the route's real path, and covers the cleanup, the reason reported, the re-application ordering, the broadcast and the caps. The mock route context gains the port method and the mux call the route needs. Refs #411 Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-09-16 11:44:37 +02:00
Randalix	a7f74f374f	fix(remote): keep the wake banner hidden after switching to a local session ... refreshHostWakeBanner clears _hostWake before calling _hostWakeTick, so the clear branch's `if (this._hostWake)` guard skipped the repaint: once the banner had appeared for an unreachable remote session it stayed up on every chat (local ones included) until a reload, and the 30s ticker never cleared it either. Render unconditionally in that branch — _renderHostWakeBanner is idempotent with a null state. Reproduced in a real browser (Puppeteer, mobile viewport): state went null but banner.hidden stayed false. Regression test added in test/host-wake-banner.test.ts (red before, green after).	2026-09-16 10:39:48 +02:00
DevvynandClaude Sonnet 5	0929694012	fix(custom-model): actually trigger the llama-swap load, not just watch for it ... Root cause of "it doesn't look like llama-swap is actually switching the model" (confirmed live: no load_model line in llama-swap's own logs after applying a selection). llama-swap has no "switch model" admin endpoint - the ONLY thing that starts a swap is a real inference request naming the model. Every previous fix (the conflict check, the loading banner) assumed a swap would start on its own; nothing ever actually asked llama-swap to load anything until the launched CLI's first real prompt did, which could be much later than "applying the selection" implied. Adds triggerLlamaSwapLoad() (custom-model-routes.ts): sends the smallest real request that will start a load - POST <baseUrl>/v1/chat/completions, max_tokens: 1, one throwaway message - fire-and-forget (never awaited by the caller; the frontend's own running-status polling is what actually confirms readiness). Wired into both apply paths (the dedicated restart route and the one-shot quick-start route), fired whenever the target model isn't already the one loaded and ready - a broader condition than the existing swapNeeded (which only gates the "this will evict another session's model" confirmation ask and deliberately stays narrow to that). modelSwapInProgress in both routes' responses now reflects this same broader condition too, so the frontend's loading banner actually correlates with a real in-flight load rather than only firing when something else happened to be loaded already. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 15:53:34 +08:00
DevvynandClaude Sonnet 5	01b32ee6cd	fix(custom-model): move the switching/loading status to a centred banner ... The "Claude started - switching to <endpoint>..." and "Loading <model> on <endpoint>... this can take a while" messages lived in the top-right toast corner along with everything else, easy to miss given they can each sit on screen for well over a minute (a real llama-swap model load). Adds _showCenterStatus() (panels-ui.js): a single, reused, screen-centred banner with a spinner, non-blocking (no backdrop, pointer-events: none on the wrapper) so it never gets in the way of using the app while it's up. Both call sites (_runCustomModelEntryViaRestart's switching message, _watchLlamaSwapLoading's loading message) now use it instead of showToast. Every OTHER status in these two flows - the llama-swap conflict warning already moved to its own modal, apply failures, cancellation, and _watchLlamaSwapLoading's own final "ready"/"still waiting" outcome - stays exactly where it was, in the corner. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 15:40:43 +08:00
DevvynandClaude Sonnet 5	2936ba6e3d	fix(custom-model): replace the native confirm() popup with an in-app modal ... The llama-swap "this will unload it for session X" warning used a native browser confirm() popup, which looks out of place next to the rest of the app's own modals. Adds #customModelSwapConfirmModal (index.html) with Cancel/Switch-anyway buttons, styled to match the app. _confirmModelSwap(message) shows it and returns a promise that resolves true/false the same way confirm() would; _resolveModelSwapConfirm(proceed) (wired to both buttons and the backdrop click) settles it. Both llama-swap conflict call sites (_quickStartWithCustomModelConfirm for the one-shot launch path, _runCustomModelEntryViaRestart for Claude's restart path) now await this instead of calling confirm() directly. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 15:29:38 +08:00
Devvyn	b1db5515d7	Merge branch 'master' of https://github.com/Ark0N/Codeman into followups	2026-09-16 15:23:52 +08:00
DevvynandClaude Sonnet 5	83033b4299	fix(custom-model): show a status toast during Claude's native-boot-then-restart window ... Claude stays on the launch-then-restart path (see runCustomModelEntry's own comment for why), but with nothing on screen during that window, a native boot that briefly talks to the cloud model read as "the endpoint didn't apply" rather than "the switch hasn't happened yet". A sticky "Claude started - switching to <endpoint>..." toast now covers the whole window from the native launch through the apply call, updated in place (never stacked) as the outcome resolves: dismissed on cancel or failure (replaced by the existing cancellation/error toast), handed off to _watchLlamaSwapLoading's own sticky toast when a model swap is in progress, or updated to the existing "Pointed at ... - restarting" message and auto-dismissed after 3s on a plain success. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 15:21:09 +08:00
DevvynandClaude Sonnet 5	f865f74a0f	feat(custom-model): launch directly on the endpoint, no restart, for 7 of 8 CLIs ... Fixes the visible double-launch reported on Codex: picking a custom-model Run-menu entry launched natively first, waited for it to settle, then restarted it in place with the endpoint applied. Necessary for the design at the time, but visibly a native boot immediately followed by a second one - worst on a CLI whose TUI fully reinitializes on a restart, confirmed live on Codex. POST /api/quick-start gains an optional customModel field ({endpointId, modelId, confirmed?}). When present, the route mints the session's id itself (crypto.randomUUID()) before constructing it, computes the same injection the existing POST /api/sessions/:id/custom-model route computes (including the llama-swap conflict check from the last commit - same {requiresConfirmation, currentlyLoadedModel, affectedSessions} shape, no session created until confirmed), and launches the session already pointed at the endpoint: env vars via the constructor, and the launchModel override merged onto piConfig/grokConfig/ompConfig using the registry's own launch.legacyConfigField the same way session.ts's restart path already does. No restart at all - setCustomModel() afterward is bookkeeping only. Wired into 7 of 8 launch functions (session-ui.js): openCode, codex, gemini, pi, grok, deepseek, omp. Claude stays on the original launch-then-restart path for now: its own --resume-based restart is far less jarring than the other seven's, and runClaude()'s multi-tab launch plus docker-config-drift confirm/retry loop make folding it into the one-shot path separate, higher-risk work than the other seven's each-a-single-simple-launch shape. Also fixes a pre-existing 'mode === omp' branch flagged by the CLI-id static guard (test/cli-registry-no-id-branching.test.ts) - the ompConfig launchModel merge is the same 'legacy <Mode>Config plumbing' category as the six sibling branches already allowlisted there, just newly literal where it was previously only inside resolveOmpConfigForCreate's own check. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 15:04:10 +08:00
DevvynandClaude Sonnet 5	fbee1b2d82	docs(changeset): add changeset for the Run-menu custom-model picker PR ... Covers #430's full scope so far: the picker itself, the model-selection dialog, periodic re-discovery, and the session-busy/toast/CLAUDE_CONFIG_DIR/ context-length/llama-swap-conflict fixes found through live validation against a real llama-swap server. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 14:39:56 +08:00
DevvynandClaude Sonnet 5	bcebc81fcd	feat(custom-model): detect llama-swap model conflicts before switching ... Root-caused the user's earlier confusion ('the terminal says opus even though something is waiting for llama to load'): llama.cpp runs exactly one model at a time, and llama-swap unloads/reloads it on demand - a swap can take anywhere from a few seconds to well over a minute, during which a session looks indistinguishable from one still on the native backend. 1. Feature-detects llama-swap (vs. plain llama.cpp/any OpenAI-compatible server) via its own GET /running, which plain llama.cpp has no concept of at all. New GET /api/model-endpoints/:id/running-status route exposes this read-only, for the frontend's polling loop below. 2. Before applying a selection, POST /api/sessions/:id/custom-model now checks what llama-swap currently has loaded. If it differs from the requested model AND another live session's own customModel selection is actively using that loaded model, the apply is refused with a {requiresConfirmation, currentlyLoadedModel, affectedSessions} payload instead of silently switching. A "confirmed: true" field on the retry skips the check. Switching with nothing else affected proceeds immediately, no confirmation asked, only ever when there is something to warn about. 3. The frontend (runCustomModelEntry) shows a native confirm() naming the affected session(s) and the model they'd lose, matching this codebase's existing convention for this class of decision (delete case, kill session, etc.) rather than a new modal. On a successful apply the response also carries modelSwapInProgress; when true, a new _watchLlamaSwapLoading poll shows a sticky "Loading <model>..." toast via the new running-status route until llama-swap reports the target model ready (bounded at 2 minutes), so a prompt sent mid-swap reads as "loading", never as silence or an answer from whatever was loaded a moment before. Checks are read-only against llama-swap's own /running - never /props, which takes a ?model= and can itself trigger a load as a side effect of asking. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 14:35:07 +08:00
Michael GrundbergandClaude Opus 5	da933d70be	feat(sessions): offer to rebuild the sessions a host reboot destroyed ... A host reboot takes the tmux server down with it, so every pane dies, reconciliation finds nothing to attach to, and the board comes up empty. Picking yesterday's work back up meant finding each conversation in history and resuming it by hand, one at a time. The boot pass now works out what the reboot killed and leaves it on offer. It runs inside restoreMuxSessions(), in the window where reconciliation has reported the dead sessions and cleanupStaleSessions() has not pruned their records yet, which is the only place the records can still be read. The board shows a banner, and nothing is created until the user clicks it. A click rather than an automatic restore is what makes the reboot heuristic acceptable. The heuristic cannot tell a reboot from a crash that took tmux down inside the same window, so it decides whether to ASK, never whether to act: a wrong yes costs a line of text the user dismisses instead of N CLI processes nobody asked for. Four things are re-checked when the click arrives rather than trusted from boot, because hours can pass and the board moves on. The owner's privilege grant re-resolves through the env clamp. The workspace must still be on disk. A conversation the user already resumed by hand from the Resume list is skipped, since two panes running --resume on one conversation would fight over the same transcript. Entries leave the plan synchronously before the first await, and the route is single-flighted, so a double-click or two devices cannot both reach the same entry. A restored session comes back attached, idle and disarmed. Respawn controllers and Ralph loops are deliberately not re-armed: a machine that just came up is the worst moment to turn an autonomous run loose. Its workspace hooks are installed by the restore route itself, because the boot-time sweep sits behind a gate that is false after a reboot and has finished long before the click; without them a session goes silently blind, with no stop or idle events for respawn, no Approvals Inbox item and no red tab on a blocking dialog. Stats collection starts the same way. The pane is new, so the conversation continues and the terminal scrollback does not. The banner says so rather than letting an empty pane read as a broken restore. The plan lives in memory only. A server restart drops it, which costs the convenience this adds and never the conversation: the conversation is the transcript under ~/.claude/projects, which the Welcome screen's Resume list and the Session Manager already read, so a dropped plan returns the user to resuming by hand. clampEnvOverridesForOwner moves to src/session-env-clamp.ts, since the question it answers is about session privilege rather than about HTTP and it now has a caller outside the route layer. Its test hook stays re-exported from session-routes.ts. Claude sessions only for this pass. The other CLIs name their thread in their own config object, which this does not thread through yet. Remote and docker sessions are skipped on purpose, because both need another host or a container to be up and a freshly booted machine cannot promise either. Refs #411 Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-09-16 08:05:55 +02:00
DevvynandClaude Sonnet 5	25f22b9839	test(custom-model): update session-custom-model route test for CLAUDE_CONFIG_DIR isolation ... Fixes the CI failure on the last two commits: this route test asserted an exact envKeys list for a claude-mode apply that predates the CLAUDE_CONFIG_DIR isolation fix, so it failed on the new CLAUDE_CONFIG_DIR entry it correctly started appending. Updates the expected list and adds assertions for the isolated config dir path and the pre-seeded .claude.json trust-approval file, matching the behavior added in the two prior commits rather than just tolerating it. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 13:23:54 +08:00
DevvynandClaude Sonnet 5	97464bfa27	fix(custom-model): pre-approve the injected API key in the isolated Claude config dir ... The CLAUDE_CONFIG_DIR isolation from the previous commit fixed the cosmetic auth warning but introduced a real regression: an otherwise-empty config directory has none of a real profile's prior custom-API-key approvals, so Claude Code stops at an interactive 'Detected a custom API key - use it?' prompt on every single launch. Confirmed live. With nobody at a TTY to answer, the prompt's own default ('No') silently refuses the very key this feature just injected, which looks like the endpoint being ignored. Adds apiKeyTrustFile to the env-kind customModelInjection capability shape ({relPath, shape: 'claude-api-key-responses'}), set on claude's entry to {relPath: '.claude.json', shape: 'claude-api-key-responses'}. The apply step merges customApiKeyResponses.approved: [apiKey] into <isolatedConfigDir>/.claude.json - the exact field a real answered prompt itself writes to (confirmed against a real ~/.claude.json after answering by hand once), so this answers the prompt in advance rather than bypassing it. Merges onto whatever the CLI already wrote into that file on an earlier launch in the same isolated directory rather than overwriting it; a missing or corrupt file is treated as empty rather than failing the apply. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 13:08:07 +08:00
DevvynandClaude Sonnet 5	0e8b1981af	fix(custom-model): isolate Claude config dir and inject real context length ... Addresses two live-validation findings on the Run-menu custom-model picker: 1. Both claude.ai and ANTHROPIC_API_KEY set warning. Claude Code still coexists an OAuth login with an injected ANTHROPIC_API_KEY in the same config directory and warns about it (confirmed cosmetic - the API key wins for actual requests, verified via a real session's own API Usage Billing line). A custom-model claude session now gets an isolated CLAUDE_CONFIG_DIR (registry-declared via a new configDirVar field, empty, no files written into it) so there is nothing to conflict with. projects is symlinked (junction on Windows) back into the real config dir so the response viewer, subagent windows and Read My Mind keep working for that session, best-effort. 2. Context-window overflow. Claude Code assumes a large default context window for a model id it doesn't recognise and never compacts, so a custom endpoint's real, much smaller context (verified live: a 400 exceeding a 16384-token llama-swap model with a stock ~33.7K-token system prompt) silently overflows. Discovery now also learns each model's real context length from llama.cpp/llama-swap's GET /props?model=<id> (n_ctx), but ONLY for a model llama-swap's own /v1/models response already marks status.value === 'loaded' - never an unloaded one, since llama-swap treats ?model= as a routing hint and probing an unloaded model risks triggering an actual, slow, GPU-swapping load as a side effect of read-only discovery. A server with no status field at all gets no enrichment rather than a guess; a model not probed this round keeps its previously-learned value until it disappears from the list entirely. Stored per model (CustomModelHost.modelContextLengths) and applied via a new contextLengthVar registry field, set to CLAUDE_CODE_MAX_CONTEXT_TOKENS for claude. Both new fields live on the existing env-kind customModelInjection capability shape, declared only on claude's registry entry - every other CLI's injection is unaffected (pinned by test). Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 12:08:26 +08:00
DevvynandClaude Sonnet 5	5c25a52f95	fix(custom-model): wait for a freshly launched session to go idle before applying ... Root cause of every 'Session is busy' apply failure reported from live testing: a just-launched CLI reports itself 'busy' for its own startup (boot spinner, workspace-trust check) well before runCustomModelEntry's apply call could reach it, and the apply route's isBusy() guard correctly cannot tell that apart from a real turn in progress — it exists precisely to refuse restarting a session mid-turn, and a fresh boot looks exactly like one from the outside. Confirmed live: replaying the identical apply call by hand against the same session, once it had settled, succeeded immediately. Fixed by waiting on the session's own readiness signal before applying: GET /api/sessions/:id/wait?until=idle&timeout=20000, one GET already built for exactly this ('Agent wait primitives', CLAUDE.md) rather than inventing a client-side poll loop. A timeout there is a normal 200 per that endpoint's own contract, never an error, so a session still busy after 20s just reaches the apply call anyway and gets the route's own honest error — now visible, since the previous commit made error toasts sticky and stopped discarding the real error text. Tests: new case in custom-model-run-menu-ui.test.ts pins the ordering (the wait call happens, and strictly before the apply call) and its exact query string. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 10:38:37 +08:00
DevvynandClaude Sonnet 5	409a6e65f9	fix(custom-model,toast): surface the real apply error, and make error toasts sticky with a close button ... Two related fixes, both needed to actually diagnose 'Session started on the native backend — could not apply the custom endpoint' reports from live testing: 1. runCustomModelEntry()'s apply call went through _apiJson(), which unwraps a success body but SWALLOWS a failure response entirely and returns null — discarding the one thing (error, errorCode) that would tell 'endpoint unreachable' apart from 'not a discovered model', 'remote/Docker session', or a dozen other real causes the apply route already reports distinctly. Switched to _api() so the actual response body is read on failure too, and the toast now includes the real message. 2. showToast() defaulted every toast, error or not, to a 3s auto-dismiss with no way to read it again — exactly what made the above generic message impossible to act on even before the fix above. Error toasts now default to sticky (duration: 0, no auto-dismiss) unless a caller opts into a duration, and every toast — sticky or not — gets an explicit close (x) button, since a sticky toast with no way to dismiss it would just accumulate across repeated failures. Tests: custom-model-run-menu-ui.test.ts's two apply tests updated for the _api() switch (their mocks previously stubbed _apiJson, which the apply call no longer goes through), plus a new test pinning that the real server error string reaches the toast on a failure. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 09:55:03 +08:00
codeman-localandClaude Opus 5	a1c35da0d8	fix(input): deliver a recovered keystroke before the Enter that submits it ... Every message typed on an Android phone lost its last character. An Android soft keyboard commits the last typed character and sends the Enter key in ONE InputConnection transaction, so the committed-text `input` event and the Enter keydown are both processed before any zero-delay timer runs. The orphaned-input recovery from #388 resolved its candidate only on such a timer, and that lost the character twice over: * ORDER — xterm emits `\r` synchronously from the Enter keydown, and the local-echo composer submits `pendingText` right there. The recovered character arrived one macrotask too late to be part of the prompt. * LOSS — that same `\r` bumps the canonical counter, so by the time the candidate resolved, `canonicalCount > snapshot` read as "xterm spoke for this keystroke" and stood the recovery down. The character was not merely late, it was dropped. Drain pending candidates synchronously at the next keydown instead, from xterm's custom key handler, which runs before xterm processes that key. The counter then still holds the value it had while the candidate's own keystroke was current, so the stand-down decision is made against the right keystroke, and the recovered byte reaches the composer ahead of whatever the new key emits. The timer stays as the fallback for a keystroke with no key after it. Physical keyboards are unaffected: there the timer has already resolved the candidate long before the next key arrives. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-09-16 09:26:30 +08:00
DevvynandClaude Sonnet 5	9a9e542a7d	fix(custom-model): bound the model-picker dialog's height and make its list scroll ... The dialog had no max-height at all, so an endpoint with many discovered models grew it past the viewport with nothing to scroll — reported live as both "takes up the full page" and "the list is truncated", which turn out to be the same bug. Gives #customModelPickModal .modal-content the same bounded-height + scrollable-body shape cronModal's .modal-lg already uses (max-height + flex column on the content, overflow-y:auto + flex:1 on the body), scoped by id rather than folded into the shared .modal-sm class three other modals already use for short, fixed content. max-height: min(70vh, 520px) scales with the viewport (a phone gets 70% of its height; a 4K display never gets a needlessly tall dialog) rather than committing to one fixed pixel value that would be wrong at either end. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 09:23:14 +08:00
DevvynandClaude Sonnet 5	5a9ff07f57	feat(custom-model): ask which model on launch when an endpoint has more than one, and re-discover models every 5 minutes ... Two enhancements requested after live-validating PR #430 against a real llama.cpp server: 1. Model picker dialog. Picking a Run-menu Custom Endpoints entry used to apply the endpoint's defaultModelId (or the first discovered model) silently. Now, via the new selectCustomModelEntry() (session-ui.js): - exactly one discovered model launches straight away, same as before - two or more open a new #customModelPickModal listing every discovered model; defaultModelId (if set) is marked but never auto-chosen, since the point of asking is letting ONE launch deliberately differ from the saved default, not just confirming it The endpoint is re-fetched at click time rather than trusting anything cached from the dropdown's own render, since the model list can have changed (the sweep below, or a settings-panel edit) since it opened. runCustomModelEntry() itself — the actual launch, routed through run() for the in-flight lock, snapshot-guarded against applying to the wrong session — is unchanged; it now just always receives an explicit model id from one of these two paths instead of computing one itself. 2. Periodic re-discovery. Every saved endpoint's models now refresh automatically every 5 minutes in the background (CUSTOM_MODEL_REDISCOVER_INTERVAL_MS, server.ts, registered the same way as the Codex plan-usage poll it sits beside — this.cleanup.setInterval, off under testMode), so a model the server starts or stops serving shows up without another manual "Discover" click. The manual POST .../discover-models route and the new refreshAllCustomModelHosts() sweep (custom-model-routes.ts) now share one pure merge step (applyDiscoveredModels: stamps lastDiscoveredAt, drops a defaultModelId that no longer appears) rather than two copies that could drift. The sweep is best-effort per host — one endpoint being unreachable on a cycle never blocks the others — and re-reads the store before each host's write, keyed by id, so a concurrent edit or delete from the settings panel always wins over a sweep that started before it. Tests: test/custom-model-endpoint-rediscovery.test.ts is a new, dedicated file for the sweep (kept separate from custom-model-routes.test.ts because that file's data dir is shared across every test in it — one temp HOME per FILE, not per test — which would make a sweep-touches-every-host assertion meaningless there). test/custom-model-run-menu-ui.test.ts gained a new describe block driving the real picker modal through JSDOM: single-model bypass, multi-model dialog with the default marked-not-chosen, picking a row closes the modal and launches with that exact model, the endpoint re-fetch, and the two "vanished by click time" toast paths. Docs: docs/custom-model-endpoints.md, docs/wiki/Custom-Model-Endpoints.md, docs/api-reference.md and CLAUDE.md's dense feature paragraph all updated — the last of these also caught up two sentences that had gone stale after the draft-review fixes landed (the picker routes through run() now, not a raw run*() call). Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 08:50:18 +08:00
DevvynandClaude Sonnet 5	60e1bd52f7	fix(custom-model): act on the draft review — unparseable onclick, unwrapped envelope, wrong-session apply, missing lock, no tests ... Addresses every blocker, both majors, and all but one minor from the maintainer's review of the draft PR. Blockers: 1. Every generated inline onclick was unparseable. JSON.stringify's own double quotes terminated the double-quoted HTML attribute at the first one, leaving btn.onclick null on every picker entry and every Discover/ Edit/Delete button. Fixed with escapeHtml(JSON.stringify(...)) per argument, the same idiom deleteCase's onclick already uses four lines away in session-ui.js. This also closes the live-HTML-injection route through modelId (server-controlled, from the endpoint's own /v1/models reply): with quoting intact, a `>` inside it can no longer terminate the <button> tag early. 2. GET /api/model-endpoints wraps its body in the {success,data} envelope like every other /api route (server.ts's preSerialization hook applies to arrays too), so Array.isArray(hosts) was always false in production and the picker/settings panel silently saw nothing. Both call sites now go through _apiJson(), which already exists for exactly this. 3. A failed or declined run*() (missing CLI, isBusy, a caught exception) returns normally without ever changing activeSessionId, so the apply step used to silently re-point and restart whatever session the user was already looking at. runCustomModelEntry() now snapshots activeSessionId before the launch and requires it to have actually changed. Majors: 4. Routes the launch through run() itself via a temporary _runMode swap (never persisted — setRunMode() would sync it to the server) instead of a parallel hardcoded dispatch table, so a custom-model launch now holds the same _runInFlight lock every other Run click gets. This also resolves the "hardcoded runners map contradicts the PR's own design" minor: dispatch is run()'s own, so a CLI whose customModelInjection recipe lands later needs no update here. 5. New test/custom-model-run-menu-ui.test.ts drives the real session-ui.js against a JSDOM window (runScripts:"dangerously" — this JSDOM only ever parses markup this module generated itself) for exactly the DOM-level facts the review said needed no Playwright and no tmux: a generated button's onclick genuinely compiles and fires, a dangerous modelId never produces a live element, the envelope unwrap works, the session-changed guard holds, run() actually gets called (proving the in-flight lock engages), and _runMode is restored afterward. Confirmed against the pre-fix code first (reproduces btn.onclick === null exactly) so this isn't a vacuous pass. Plus new tests in custom-model-routes.test.ts and render-index-html.test.ts for the other fixes below. Minors: - Generated entries now filter through isCliAvailable(), matching _refreshRunModeAvailability's own gating of the stock entries. - The CRUD panel is now gated on customModelEndpointsEnabled (applyCustomModelEndpointsVisibility(), wired to the toggle's onchange and to settings-modal open) instead of always rendering; the endpoint GET no longer fires unconditionally either. - API keys are never handed back to the browser on GET, POST or PUT — redactApiKey() replaces the field with a computed apiKeySet: boolean, and a PUT with no apiKey now keeps the stored one server-side (applyStoredApiKey()) instead of the client resending a value it was never given. New tests cover both directions (kept vs. replaced) by observing the actual auth header a subsequent discovery request sends. - "+ Add endpoint" hides for a non-admin in multi-user mode (_applyCustomModelAdminGate(), also wired to admin-ui.js's codeman:me event, since the real role can resolve after settings were first opened) — endpoint writes were already admin-only server-side, but the button used to render for everyone and eat a 403. - design doc (custom-model-endpoints-plan.md §4) now says up front that its toolbar-button design was superseded by the Run-menu picker. - docs/api-reference.md gained a Custom Model Endpoints section (every route, the apiKeySet/defaultModelId contract, the restart mechanics). - Wiki page now covers un-pointing a session (curl/delete, no UI yet) and that the picker is desktop-only for now. - .set-inline-form uses --control-bg instead of a hardcoded black alpha (CLAUDE.md already records that exact literal turning the settings preview into a grey slab on light skins), .run-mode-custom-models gets the same gap: 2px .run-mode-menu's own flex gap only applies one level up, and the index.html comment naming the wrong function is fixed. - __codemanCustomModelClis's JSON is now escaped against a literal </script> (CliEntry.label is user-clis.json-settable, unlike __codemanCliAvailable's booleans-only payload) via a new exported escapeScriptJson(), pure and unit-tested without needing a WebServer. - Added defaultModelId + the new /v1/model-endpoints routes to docs/api-reference.md; left the "no zh-CN for the new Models-section group" minor unaddressed only insofar as the wider Models section (task routing, thinking effort, etc.) has never had zh-CN coverage either — everything this PR itself introduces (labels, hints, button text, the Run-menu's "Custom Endpoints" header) IS translated in i18n.js. Regression caught while fixing #4: the admin-gate's codeman:me listener is a module-level document.addEventListener() call, which threw in run-mode-ui.test.ts's minimal vm-context fake document and failed all 10 of that file's tests. Fixed with optional chaining before it ever reached the branch this commit lands on; full targeted suite (route tests, structural guards, every settings-ui.js-loading frontend test) reverified green afterward. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 07:01:23 +08:00
DevvynandClaude Sonnet 5	fed6582d3e	fix(test): strip the custom-model Run-menu picker's injected script too ... CI on PR #430 failed test/server-index-title.test.ts's byte-identity check: renderIndexHtml now injects a second unconditional <script> before </head> (window.__codemanCustomModelClis, added alongside the existing __codemanCliAvailable one), and the test only knew to strip the older one before comparing the rendered HTML against the raw template. Strip both. Unlike __codemanCliAvailable (an object, historically injected only where something resolved), the new one is a plain array injected unconditionally, possibly empty, so it needs stripping on every machine, not just one with CLIs installed. Verified the two replace() calls compose correctly against the exact strings server.ts actually produces (simulated in isolation; this box has no tmux, so the real WebServer-backed test file cannot run here at all -- same environment gap noted throughout this PR's review). Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 07:01:23 +08:00
DevvynandClaude Sonnet 5	98d26e14d9	docs(wiki): document Custom Model Endpoints and the Run-menu picker ... New docs/wiki/Custom-Model-Endpoints.md (auto-synced to the live GitHub wiki on push to master, per docs/wiki/Contributing.md) covers turning the feature on, adding an endpoint, the Run-menu picker's one-off-run behaviour, the per-harness confidence table, and what it deliberately does not do yet (remote/Docker sessions, live hot-swap). Linked from the sidebar, from Agent-CLIs.md's "Read next" list plus a short pointer section, and from Settings-Reference.md's Models section. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 07:01:23 +08:00
DevvynandClaude Sonnet 5	25fae9ad10	feat(custom-model): generate Run-menu entries from saved endpoint profiles ... Follow-up to #393, picking up the work Ark0N invited in his merge comment: "generate those entries from the saved profiles rather than a fixed duplicate per harness, and put it in a follow-up PR so this one stays the backend... The Run-menu picker is yours if you want it." Adds the frontend surface the backend has been waiting on: - Run menu: a "Custom Endpoints" section lists one entry per (harness that supports customModelInjection, saved endpoint) pair, e.g. "Claude Code (llama.cpp)". The harness list comes from window.__codemanCustomModelClis, injected at page render straight off the CLI registry's own capabilities (never a hardcoded id list in the frontend), so a CLI whose injection recipe lands later appears with no frontend change. Picking an entry runs that harness's own existing run*() function unmodified (case creation, env overrides, everything, forced to a single instance) and then applies the endpoint's default model to the session it creates via the existing POST /api/sessions/:id/custom-model route. Entries are hidden for a remote/docker active case, since that route already refuses both. - Settings: App Settings -> Models gets a "Custom model endpoints" group wiring up the customModelEndpointsEnabled toggle (declared since #393, read by nothing until now) plus CRUD against the existing /api/model-endpoints routes: list, add/edit (inline form), delete, discover models. - Backend: CustomModelHost gains an optional defaultModelId, the model the picker applies with no further choice per endpoint (one generated menu entry per CLI+endpoint pair, not per CLI+endpoint+model). The route refuses a value that isn't one of the endpoint's own discovered models, and a fresh discovery drops a default that no longer appears rather than carrying an invalid one forward. Docs: docs/custom-model-endpoints.md describes the new picker and settings panel; CLAUDE.md's Custom Model Endpoint Profiles entry drops the "backend-only" status note and documents the picker's generation mechanism. Tests: four new route tests cover defaultModelId validation, acceptance, and the drop/keep behaviour across a re-discovery; a new render-index-html test pins the __codemanCustomModelClis injection (present, agent CLIs supporting the capability, antigravity and shell excluded) and its solo-window skip. No browser test was added for the Run-menu picker itself or the settings CRUD panel (this box has no tmux, so the live server used by test:browser/test:mobile could not be exercised here) -- worth a Playwright pass before merge, same as any other frontend PR. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01RqZeHrRS6DYcGcGX2p9EwG	2026-09-16 07:01:23 +08:00
Randalix	4a30f510e6	fix(remote): let the host config turn wake-on-LAN OFF for a live session too ... Found by driving the real UI: with a MAC configured in remote-hosts.json, removing it (here: to reach the "Configure WoL" dialog) changed nothing for a running session — _effectiveRemote short-circuited on the session's own snapshot whenever that snapshot HAD a target, so the resolver was only ever consulted in the one direction where the feature was missing. The documented "host config is authoritative" promise therefore failed in the direction a user can actually observe, and a wake target could live on invisibly after being deleted from the config. The resolver is now consulted on the TTL regardless, and wins for the wake fields in both directions. Also adds a route test for the browser's real input shape: one POST per keystroke, all buffered during a wake, replayed IN ORDER.	2026-09-15 23:21:50 +02:00
Randalix	d0a5a583cd	feat(remote): wake a sleeping host when a session is created or attached ... Pressing Run on a remote case whose host was asleep failed with `could not verify tmux on remote host 192.168.50.137: …` — an ssh error that blames tmux for a machine that is merely suspended. The only wake paths were typed input on an established session and the banner's Wake button, so OPENING a session (the moment the user actually decides to use that host) had none. `RemoteWakeRegistry.ensureHostAwake()` reuses the existing probe/wake/readiness machinery for a host that has no session yet, and is wired into the two user-initiated create paths: `POST /api/quick-start` for a remote case (before the tmux prereq probe, which is what surfaced the misleading error) and `POST /api/sessions` with `attachRemoteSession`. A host without a wake target is not even probed, so its behavior and latency are byte-identical. The wake is blocking — the caller gets the session or an error — but bounded by REMOTE_WAKE_REQUEST_READY_TIMEOUT_MS (40 s) instead of the 90 s session default, because the dashboard sits behind a reverse proxy whose default `proxy_read_timeout` is 60 s: a longer wait would be cut off at the proxy while the session was still being created. The budget has to cover the whole request (40 s wake + 1.5 s probe + the tmux probe's own 15 s = 56.5 s worst case), which is why it is 40 s and not 45. A timeout now says the host did not come back, and an unreachable host without a wake target says so instead of pointing at tmux. The wiring is deliberately in the HTTP ROUTE, never in the shared session service: `cron-service.ts` builds sessions there with nobody waiting on the answer, and a wake on that path would power the host on for every schedule — the timer-driven re-wake invariant #1 exists to prevent. Both halves are asserted (importers of `remote-wake`, and `ensureHostAwake` having exactly one caller file), so a future caller has to come through the guard test. A rejection from the wake IO is caught too: a broken target must fail the wake, not the route. `remote:hostWaking`/`remote:hostWakeFailed` now carry `forNewSession` for the session-less case, where "input is queued" would be untrue; the toast then reads "the session starts when it is back". Live wake numbers are unchanged (this reuses the measured ~12 s S3 path); the route behavior is covered by new tests in session-routes.test.ts with an injected registry, so no test opens a real socket or ssh.	2026-09-15 22:37:37 +02:00
Randalix	8dfc965d13	fix(remote): stop the wake handlers shadowing each other; enforce the input cap ... Two findings from a final review pass over the wake-on-LAN feature. `_onRemoteHostWaking` / `_onRemoteHostWakeFailed` were defined in BOTH `panels-ui.js` (toasts) and `host-wake-ui.js` (banner). Both files mix into `CodemanApp.prototype` and `host-wake-ui.js` loads later, so the panels-ui copies were silently shadowed: the toast never fired, and a wake started for a BACKGROUND session (input on a non-active tab) produced no notification at all, since the banner handler only acts on the active session. The handlers now live only in `host-wake-ui.js`, show the toast unconditionally, and update the banner when the woken session is the active one. `appendBoundedPending` dropped only WHOLE chunks, so a single input value over the cap (one large paste is one `input` value, up to the 100 KB input schema) was kept in full: "bounded at 4 KB" held per chunk, not per session, and nothing was logged. The surviving chunk's head is now trimmed too, code-point aware so a multi-byte character is never split into a replacement char. Adds the guard that would have caught the first one: every SSE dispatch handler must be defined in exactly ONE frontend module. The existing test only asserts a handler EXISTS somewhere, which two modules both satisfy while one is shadowed.	2026-09-15 21:01:53 +02:00
Codeman maintainer	bd286bf502	docs(wiki): catch the manual up to 1.29.0 and add the three run modes it never had ... The wiki was written for seven run modes and never received Grok Build, DeepSeek Harness or OMP. They now appear everywhere the others do: the modes table and per-CLI notes, install commands, environment prefixes, the Quick Start table, the requirements rows, the vocabulary, and every "seven modes" count. The 1.27 to 1.29.0 changes land on the pages that own them: attaching a case to an existing container, multi-case adoption and the copy-a-case picker (Docker Cases); file reads over ssh in remote cases and what stays unavailable (Remote SSH Sessions, Working With Files, Security); single-page app routing, frame recovery, localhost links as tabs and the egress guard (Web Tabs); DeepSeek as the one non-Claude mode with real stop/blocked signals and Approvals items, Codex's own work detection, last-response, the model-endpoint routes and refreshed counts (HTTP API, Driving From An Agent, Hooks, Notifications, Keeping Agents Running, Core Concepts); Shift+drag, right-click copy, Auto Copy, the Ctrl+Z guard, font weight, the vertical rail and its activity sort (Keyboard Shortcuts, Input And Voice, The Dashboard, Settings Reference); the 600px phone cutoff, Codex shift arrows and iPhone Duo (Mobile Guide); the Docker Compose route and its update rule (Installation, Running As A Service); four new symptom entries and a "which CLIs" question (Troubleshooting, FAQ). Custom model endpoints are deliberately left to #430, which adds that page and edits Agent CLIs, Settings Reference and the sidebar; these edits stay out of the regions #430, #428 and #376 touch, and all three still merge cleanly on top. Both READMEs: the web-tab menu entry is labelled "Add URL" in the UI, not "Add dashboard". Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>	2026-09-15 19:05:59 +02:00
3 people	3248f35081	chore: version packages (#437) ... * chore: version packages * chore: sync the CLAUDE.md version line to 1.29.1 Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> --------- Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: Codeman maintainer <noreply@anthropic.com> codeman@1.29.1	2026-09-15 18:18:09 +02:00
Michael GrundbergandClaude Opus 5	c9515b1d4c	fix(terminal): keep the output a pane capture could not contain ... Live terminal events are queued while a buffer load runs, and the load discards that queue when it ends. That is right when the loaded buffer is the server's accumulated byte history. The route appends to that history right up to the moment it serializes the response, so a queued event already appears in it and replaying it would duplicate output, most visibly Ink's cursor-up redraws. A tmux pane capture is a photograph, current only as of the instant `capture-pane` ran. Output printed afterwards was queued and then dropped, and nothing scheduled a re-fetch to recover it: `_onSessionNeedsRefresh` is wired only to the 128KB overflow path. The CLI's next partial redraw then landed on a frame the terminal never received. How much went missing depended on which capture the route served. A `?full=1` load returns the capture alone, with no history in front of it, so it lost everything from the capture to the end of the chunked write. A `?tail=` load returns history, a clear, and then the capture, and the route reads that history after the capture, so it lost everything from the response to the end of that write. The chunked write dominates either way. An agent CLI hides the loss on its next full redraw; a shell session does not, because its output is linear and nothing repaints it. Queue entries now carry their arrival time, and `_finishBufferLoad` takes a `since` cutoff, so a capture load replays exactly the tail that arrived after the response headers. The earlier events stay dropped, because a payload that carries history does hold those. All four paths that fetch a terminal buffer and write it now decide this the same way, through one `_bufferLoadFinishOpts` helper, so they cannot drift apart: `selectSession`, `_onSessionNeedsRefresh`, `_onSessionClearTerminal` and `_maybeRefetchFullHistory`. The second of those is the one that stings. It exists to restore output the client already dropped once under backpressure, and it was dropping more output while performing that recovery. The cache-hit write inside `selectSession` stays on discard deliberately: it runs before the fetch, so its queue holds only events the capture that follows already contains. Two further things had to change for that tail to still exist when the load ends, and a browser test is what found both. `chunkedTerminalWrite` is what ends the load for every non-empty buffer, so the flush policy travels to its own finish calls; the call in `selectSession` runs only when the write was skipped. `_beginBufferLoad` no longer empties the queue when one load re-enters it, which it does on every write, because that reset discarded the whole fetch window before anything could replay it. The response already distinguishes the sources. `source` reads `mux-visible` or `mux-full-history` for a capture and `history` for the byte stream. Follows #395, #396 and #397, which fixed the ways the replayed frame itself could disagree with the terminal. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>	2026-09-15 18:14:18 +02:00
Codeman maintainer	5b920cb43d	feat(sessions): land auto-naming opt-in, in the prefix form, from the first user prompt only ... Finishes #376. The contributed keystroke tracker sat on the raw byte stream and named tabs wrong five ways (every prompt, every write path, a bare Esc eating the next prompt's first character, pasted newlines as Enter, any CSI clearing the draft) and replaced the whole name, which dropped the case from the tab and reset the w<n> counter. This lands the feature with each of those closed: - First prompt means the first: applyAutoName() flips a placeholder to `auto` whether or not the string changed. nameSource is now the tri-state placeholder | auto | manual; the name setter is the only manual path. - Only user-originated input counts: write()/writeViaMux() take SessionWriteOptions.fromUser, set by the browser WS path and POST /input only, so Ralph, respawn, cron, approvals and the trust-dialog keys can never name a tab. A startMode 'shell' CLI never feeds the tracker (a capability, not an id check); the send-key route feeds trackUserInput() because its line feed bypasses the session. - Prefix form `w3-case: title`: parseSessionPrefix() already renders it as the title with the prefix in the tooltip and the next-session counter still matches it. Composed within MAX_SESSION_NAME_LENGTH. - Tracker rules per key: bare Esc resolves at chunk end; mouse/focus reports, Tab, cursor keys, Shift+Tab are no-ops; Up/Down and Ctrl+P/N/R taint the draft so Enter submits nothing rather than a fragment; bracketed-paste newlines and Ctrl+J / Shift+Enter join with one space; the draft keeps its head past 8192 code points; an escape past 64 bytes is abandoned. - Title: slash commands by shape (a path is a prompt), `!` escapes refused, first sentence only past 8 code points ("e.g." is not a title), 72 code points on a word boundary. - Synced `autoNameSessions` setting, default OFF (the prompt reaches mux-sessions.json, session:updated and /api/search), App Settings -> Appearance -> Tabs, read fresh per prompt after the eligibility check. Tests: test/session-auto-name.test.ts (tracker, title, composition, ownership, emit gating), the wiring test (once, prefix, setting off, manual protected), test/routes/session-name-routes.test.ts (PUT /name flips to manual and persists). Verified live on an isolated instance: API and browser-typed prompts name the tab, a second prompt does not, shells and renamed tabs are untouched, nameSource survives a restart. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>	2026-09-15 17:59:16 +02:00
Codeman maintainer	c4322513d9	Merge pull request #376 from shenlvkang-collab/feat/auto-session-names-upstream	2026-09-15 17:19:57 +02:00
Randalix	1380b023e2	fix(remote): make the wake banner's poller page-wide and independent of tab switches ... Reported as 'the tab shows no banner' while the host was verifiably unreachable: the banner only started polling from selectSession, which RETURNS EARLY for the tab you are already on (so a page loaded with the remote tab active never polled), and a long-lived tab keeps running the JS it loaded — the feature was invisible to anyone who did not switch tabs after the deploy. The poller is now page-wide: one interval (created on init and on the first session switch), re-targeted whenever the active session changes, plus a visibilitychange wake-up. It no longer depends on any single selection path running. Also adds test/sse-dispatch-table.test.ts: a static guard that every [SSE_EVENTS.X, '_onFoo'] entry names an event constants.js defines AND a handler some module defines. Both halves fail silently (a typo'd constant is an undefined table key; a renamed handler just never runs), which is exactly how a new banner can never appear with no error anywhere.	2026-09-15 15:11:36 +02:00
Randalix	e8f7772320	fix(remote): offer the WoL config dialog after a failed wake too ... A configured-but-broken target (host replaced NIC, command removed) had no way out: the dialog hung off the 'no target configured' branch only, so the banner would keep offering a Wake button that keeps failing.	2026-09-15 14:38:24 +02:00
Randalix	2f61be6e74	fix(remote): bind the wake socket before enabling broadcast ... setBroadcast() on an unbound dgram socket throws EBADF on Linux and the following send fails with EACCES, so the magic packet silently never left the machine — the feature reported a wake that never happened. Caught by waking a real sleeping host (a unit test with a real UDP broadcast would not be welcome in CI, so the socket is injectable and the bind-before-setBroadcast ORDER is asserted).	2026-09-15 14:24:13 +02:00
Randalix	8b5a13435a	feat(remote): host-unreachable banner, manual wake, and native MAC wake-on-LAN ... The reactive wake (typing into a session whose host slept) left the state invisible: nothing told the user the machine was asleep, and with no wake target configured there was nothing to do about it. Adds: - RemoteHost.wakeMac (comma-separated) - Codeman builds and broadcasts the magic packet itself (UDP port 9), so the common case needs no external script. The existing wakeCommand stays as the explicit override. - GET /api/sessions/:id/reachability - probes (throttled, cached, and it never wakes) and reports HOW the host can be woken, or that nothing is configured. - POST /api/sessions/:id/wake - wakes, waits, reattaches the pane and flushes buffered input; 400 with a routable message when no target is configured. - The amber host-unreachable banner + its 'Wake' / 'Configure WoL' action, and a small config dialog that saves via PUT /api/remote-hosts/:id. - RemoteWakeDeps.resolveRemote: host config is re-resolved for LIVE sessions (throttled + cached), so saving the dialog takes effect without a restart.	2026-09-15 14:20:31 +02:00
Randalix	3f0bfde54a	docs(remote): document the wake-on-LAN invariants; drop wake state on bulk delete ... Self-review pass: the input-ladder's two 'buffer' branches were the same three lines, and bulk delete left a session's (bounded, per-random-uuid) wake state behind. Documents the design where the code refers to it - remote-sessions.md section, the architecture invariant, and the CLAUDE.md key pattern.	2026-09-15 10:45:01 +02:00
Randalix	0f3eea2fb5	fix(remote): refresh wake command from host config when restoring sessions ... A session's remote block is persisted at launch time and recovery uses that snapshot, so a wakeCommand added to remote-hosts.json afterwards never reached an already-running session - not even across a Codeman restart (observed: the live Hufflepuff session came back with no wakeCommand). Merge the host-level field in on restore, with the host config authoritative.	2026-09-15 10:35:58 +02:00
Randalix	a81f430e41	feat(remote): wake a sleeping host from user input (Wake-on-LAN) ... A durable remote session survives SSH drops (COD-104/108), but nothing brought the HOST back: after the remote machine suspended, the local tmux pane's ssh child stalled silently and `send-keys` SUCCEEDS against it, so typed input vanished with no error anywhere. Add an optional per-host `wakeCommand` (Wake-on-LAN wrapper, e.g. whuff) that the input route runs when a wake-enabled host is unreachable: input is buffered, the host is woken, the pane is reattached, and the buffer is flushed in order. Detection is a throttled bare TCP probe on wake-enabled hosts only, and only REAL user input may wake a host - the auto-reconnect watcher and boot recovery deliberately cannot, or the host would be re-woken seconds after every suspend and could never stay asleep.	2026-09-15 10:25:52 +02:00
DevvynandClaude Sonnet 5	3f2928ae73	chore(cli-registry): clean up dead code and stale claims left after #380 ... Addresses the "left as they are"/"worth knowing" items Ark0N named when merging #380 (the CLI-catalogue-driven install.sh + Docker agent image PR), none of which were correctness-blocking but all of which were real: - Removed install.sh's dead _cli_index/check_cli/get_cli_path helpers: the catalogue-driven menu and hints stopped calling them and nothing else ever did. - The generator no longer emits CLI_KIND/CLI_NPM, two bash arrays install.sh never read (the .mjs/docker-hosts.ts producers already read the JSON catalogue's kind/npmPackage fields directly, so only the bash copies were dead). - detect_all_clis now skips a disabled entry's probe entirely instead of running it and filtering the result downstream. No stock entry ships disabled today, so this closes a latent inefficiency before it is a latent bug rather than fixing an observed one. - The install hint for a launcherProfile entry (DeepSeek today) now explains in one line why it's a docs link and not a command: its own docs page documents `npm install -g @deepseek-ai/dsh`, which installs the launcher only and can't drive a pane, the exact trap the menu already avoids by withholding the command. Driven by a new generated CLI_LAUNCHER_ONLY array (from discovery.launcherProfile), not an id check, so any future launcherProfile entry gets the same caveat free. - Corrected the non-interactive-default comment: on a wget-only host, Claude's curl one-liner is filtered out of the offered list first, so the default becomes whichever npm-based entry sorts earliest instead (Codex today), not always Claude. Behaviour is unchanged — it was already printed, never silent — only the comment overclaimed. Tests: extended test/install-sh-invariants.test.ts with a positive guard for the new array and the trimmed array list, a negative guard that CLI_KIND/CLI_NPM/the three dead helpers cannot come back, and two real-bash tests (driven the same way the existing skip-menu tests are) proving a disabled entry is genuinely never probed rather than merely filtered after the fact. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_011WzDjJnbK7zug8iQWnCc9z	2026-09-15 09:10:37 +08:00