Agent working notes (HANDOVER/handoff) and one-off UI prototype HTML files were committed by mistake. Remove the already-tracked ones, add .gitignore patterns for these classes of files and a .tmp/ scratch dir, and document the rule plus a pre-commit self-check in AGENTS.md so future mistakes are caught mechanically.
No publishable package is affected, so this PR needs no changeset.
* feat(web): render AgentSwarm as an inline tool card
Replace the bottom SwarmCard footer and the messagesToTurns live-skip
with one dedicated inline tool card for AgentSwarm. The card shows a
phase overview plus a per-subagent accordion: live progress while it
runs, parsed aggregated result once it completes (and after a refresh
that has already dropped the live tasks).
Refresh and resync keep member identity metadata (swarmIndex,
parentToolCallId, subagentType, runInBackground) stable across skeleton
task replacement in the reducer, and the .content-wrap flex layout is
hardened against the overflow compression that previously displaced the
footer.
* fix(web): handle swarm review feedback
- SwarmTool: when AgentSwarm fails before producing a structured
agent_swarm_result (e.g. argument validation), render the raw tool
output instead of the "waiting for subagents" placeholder so the
failure cause is visible.
- resolveSwarmMembers: source live members from the AppTask store keyed
by parentToolCallId rather than buildSwarmGroups, which filters out
single-member groups. A resume-only AgentSwarm now streams its live
progress before the final result arrives. The badge counter still
relies on buildSwarmGroups's filter.
* fix(web): carry streamed subagent text into swarm rows
Swarm subagents that stream normal assistant output accumulate it on
AppTask.text (text-kind taskProgress), not outputLines. The new live
member map was dropping `text`, so a still-composing subagent rendered
an empty / stale row until the structured result arrived.
- Add `text` to SwarmMember and thread it through buildSwarmGroups and
swarmMembersByToolCall.
- SwarmTool: prefer member.text for both the row activity preview and
the expanded body; fall back to outputLines / summary.
- Tests cover text propagation through both helpers.
* fix(web): merge swarm result rows and fall back to raw output
Address the two latest swarm review comments:
- Rows: when a parsed agent_swarm_result coexists with live AppTasks
(which the detail panel also depends on), the inline card previously
only rendered the live members. Interrupted swarms can carry
state="not_started" / outcome="aborted" result entries for items that
never spawned a task; those rows were dropped until a refresh cleared
the live tasks. Extract the row model into buildSwarmCardRows and
merge result-only aborted/not-started rows with the live member rows.
- Fallback: when the tool is no longer running but produced no
structured result (argument validation, parser miss, or legacy
legacy transcript), render the raw tool output instead of
"Waiting for subagents…" so the final text / failure cause is
visible to the user.
* fix(web): parse swarm result subagent bodies defensively
Producer writes subagent body unescaped, so a subagent that analyzes or
emits an AgentSwarm snippet can include a literal "</subagent>" inside
its body. The non-greedy regex treated that as the row close and truncated
the body in the result-only path (post-refresh where the AppTask store is
gone).
Rewrite the parser to scan opening tags, then resolve each row's body as
everything up to the last "</subagent>" before the next row's opening tag
(or document end), preserving embedded close-tag strings. Add tests for a
literal "</subagent>" within a single body and across sibling rows.
* fix(web): only count top-level subagent result tags
A subagent body that contains a literal `<subagent ...>` tag — for
example emitting an AgentSwarm/XML snippet — was being pre-collected as
another result row, splitting the real body and producing duplicate /
bogus subagents after refresh where the AppTask store is gone.
Rewrite parseSubagents with a depth-tracking tokenizer: scan every
`<subagent ...>` / `</subagent>` token in order, push a real row frame
only at the outermost level, and treat openings / closings while nested
inside another body as body text. Drop the now-inaccurate "literal
</subagent> without matching open" regression tests; replace with tests
that verify a balanced nested snippet stays inside the parent body and
does not register as a separate row.
The desktop composer toolbar renders every control on one row and relies on
overflow:hidden to fit, so between the mobile breakpoint and a wide window it
clipped its own content. Shed secondary ink below 980px (the context readout
moves into the ring tooltip, the model name truncates earlier, permission is
capped) and keep the context ring visible on phones too, instead of sending it
to the settings sheet.
* feat(telemetry): add system metrics collection
Add periodic CPU and memory telemetry sampling with warmup capture, lifecycle cleanup, and tests.
* fix(telemetry): attach prompt session to system metrics
* fix(web): reconcile session from snapshot on reopen
* fix(web): discard stale snapshot when a newer prompt races reopen
* fix(web): harden reopen snapshot against first-open and optimistic-send races
* fix(web): keep evicted reopens subscribed when a snapshot races
* fix(web): let resync snapshots bypass the reopen staleness guard
* fix(web): force-apply the snapshot after an undo
* fix(web): gate session reopen on durable seq instead of updatedAt
* refactor(web): always rebuild reopened sessions from a snapshot
* fix(web): sharpen reopen snapshot discard and skip rebuilds mid-stream
* refactor(web): unconditionally apply session snapshots, drop the staleness guard
* fix(web): preserve loaded older messages when reopen snapshots apply
Add an `interrupt_reason` field to the `turn_interrupted` telemetry event so the data can tell a deliberate user cancel (`user_cancelled`) apart from a programmatic abort (`aborted`), max-steps exhaustion (`max_steps`), an error (`error`), or a hook-filtered turn (`filtered`).
The user-cancel signal comes from the existing UserCancellationError carried as the abort signal's reason, reused here without changing any loop control or external protocol semantics.
* feat(server): support restoring and listing archived sessions
- add a `:restore` session action that clears the archived flag in state.json and returns the restored session
- add an `archived_only` list query param, mutually exclusive with `include_archive`, that post-filters to archived sessions
- keep the implementation in the server layer as a temporary measure until agent-core exposes restore natively
* fix(server): paginate archived-only sessions before response
* feat(web): add archived sessions page in Settings
Browse, search, filter by workspace, sort, and restore archived sessions
from a new Archived tab in Settings, backed by the server archived_only
list and :restore action.
* fix(web): keep archived Load more visible when a page filters to empty
When a search or workspace filter empties the loaded archived page, the
Load more button was hidden inside the non-empty branch, so users could
not fetch older pages to find a match. Move the button out so it stays
available whenever more archived pages exist.
* fix(server): preserve after_id bound while draining archived pages
Draining an archived_only request that starts from after_id would switch
to before_id and cross the pivot, reintroducing the pivot and older
sessions. Take a single filtered page for after_id instead of draining
past the lower bound.
* fix(server): drain archived_only within the after_id bound
An archived_only request starting from after_id now keeps paging toward
older sessions until it reaches the pivot, instead of treating the first
page as exhaustive. The loop stops as soon as it encounters the pivot
session itself, so it never reintroduces the pivot or anything older.
* feat(web): drain all archived pages for global search and sort
When the user searches, sorts, or changes the workspace filter in the
Archived settings page, fetch every remaining archived page first so the
client-side filter and sort run over the full set rather than only the
pages loaded so far.
* refactor(web): load all archived sessions upfront instead of paginating
Fetch every archived session once when the Archived settings tab opens and
drop frontend pagination entirely. Search, sort and workspace filter now run
over the full set, removing the empty-page and cursor bookkeeping that
previously caused bugs.
---------
Co-authored-by: qer <wbxl2000@outlook.com>
* fix(kimi-web): keep composer caret visible when input is empty
The composer textarea coloured its empty state with `--faint`, and with no caret-color set the caret inherited that faint colour and nearly vanished until the first character was typed. Pin the caret to --color-text so it stays readable regardless of the placeholder state.
* fix(kimi-web): use currentColor for done todo strikethrough
TodoCard pinned the done-state strikethrough to --color-line-strong, which is lighter than the faint text in light theme and darker in dark theme, so the line looked washed-out and broken against the text. Drop the override so the line inherits the text colour, matching TasksPane and the design-system examples.
* fix(kimi-web): do not reserve width for hidden workspace row actions
The workspace header's more/add buttons were hidden with opacity: 0, which kept them in the flex layout and permanently reserved ~60px on the right, so the workspace name (flex:1) truncated long before the row was full. Hide them with display: none and restore display: inline-flex on hover/focus/open, so the name fills the row and only truncates once the buttons actually appear.
The same actions remain reachable via the right-click menu and the section kebab, so removing them from the tab order when hidden is acceptable.
* fix(kimi-web): keep workspace header row height stable on hover
Lock .gh-top to the sm IconButton height (26px) so revealing the hover actions no longer grows the row and nudges the path line and the groups below it. Follow-up to the display: none change: that freed the horizontal space but let the row height collapse when the buttons were hidden.
* chore: add changeset for kimi-web UI fixes
* fix(kimi-web): restore keyboard access to workspace actions
Revert the display: none change (and the min-height that accompanied it) for the workspace row's hover actions, going back to opacity-hidden buttons.
The display: none approach removed the buttons from the tab order, and the 'create in workspace' add button has no keyboard-accessible alternative in the right-click or section menus, so keyboard users lost that action. Restoring opacity keeps the buttons focusable again, at the cost of the workspace name truncating a little earlier to reserve their space.
Addresses the P2 review on PR #1423.
Keep the command preview in the body after the result lands so a multi-line command with short output no longer collapses the card. Render the command in textDim with a shellMode $ and the result one shade dimmer in textMuted, and simplify the header to "Running a command" / "Ran a command".
* feat(agent-core): feed AskUserQuestion answers back as question text and option labels
The flattened answers record the model receives was keyed by synthesized
ids (q_0 / opt_0_1), forcing a cross-message positional lookup against the
original tool call to understand what the user picked — both unreadable in
transcripts and a real model-misreads-the-choice badcase.
- toAgentCoreResponse now takes the original broker request and translates
wire ids back to question text (keys) and option labels (values);
unknown ids are kept verbatim, missing request falls back to raw ids
- wire protocol unchanged: clients still answer with option ids; the
resolve route reads the pending request before settling it
- question texts must be unique per call and option labels unique per
question, enforced in the tool execution path (AJV cannot express the
zod refine) and mirrored on the exported schemas
- web transcript card resolves both the new label form and legacy id
transcripts; TUI and ACP paths already produced the text form
* fix(agent-core): align multi-select answer join across clients and harden question schema
- Join multi-select labels with ', ' in the server translator, matching
what the TUI reverse-RPC path already emits, so the model sees one
format regardless of which client answered
- Trim segments in the web transcript resolver before label matching:
TUI-answered multi-select transcripts (', '-joined) previously lost
their highlight to a spurious leading-space Other row
- Move the question-text/legacy-q_<i> answer lookup out of the SFC into
askUserToolParse as answerFor(), per that module's testability intent
- Require non-empty question text and option labels (.min(1)) so empty
strings are rejected by AJV at the tool boundary instead of failing
deeper in the protocol layer
* fix(agent-core): resolve option ids only within the answered question
The translator's option-id lookup was a single flat map across all
questions, so a stale or malformed response pairing one question with
another question's option id (q_1 + opt_0_0) was silently translated
into a label that was never offered for that question. Scope the lookup
to the answered question's own options; cross-question and unknown ids
now both pass through verbatim, staying diagnosable.
Some agent tooling looks for a CLAUDE.md at the repo root. Add a
symlink to the canonical AGENTS.md so both filename conventions
resolve to the same instructions, avoiding any duplicated content.
* feat(agent-core): progressive tool disclosure via select_tools
Keep MCP tool schemas out of the immutable top-level tools[] and let the
model load them on demand, preserving the provider prompt cache:
- kosong: Message.tools (append-only load primitive, serialized as Kimi
messages[].tools with type:function wrapping and no content),
Tool.deferred (stripped once in generate() so loaded tools stay
executable without re-entering the top level), select_tools capability
bit (UNKNOWN/catalog default false).
- select_tools builtin: load-by-exact-name, three-branch semantics
settled per name (Loaded / Already available / Unknown), schemas read
from the live registry, injection-origin schema messages survive undo.
- ToolsDiffInjector: <tools_added>/<tools_removed> announcements at turn
boundaries and post-compaction, folded from history (undo/compaction/
resume self-heal), appended only when the loadable set changes.
- Loaded-tools ledger = history scan + defer-window pending set (cleared
on /clear); loop re-reads the executable table per step so a selected
tool dispatches on the next step of the same turn; preflight
distinguishes not-loaded from loaded-but-disconnected.
- Cross-cuts: projection strips protocol context for non-select_tools
models (lossless mid-session model switch both ways), compaction
filters it from the summarizer input and rebuilds loaded schemas
keep-all after folding, token estimation counts message.tools, request
logging reflects the post-strip wire tools.
- Three-condition gate: capability.select_tools x capability.tool_use x
tool-select experimental flag (KIMI_CODE_EXPERIMENTAL_TOOL_SELECT).
Any gate closed reproduces the inline request byte-for-byte; all
current models keep the capability off, so behavior is unchanged
until a supporting model is catalogued. The SDK catalog-to-alias
mapping forwards the capability so catalog-driven setups can enable it.
* feat(kosong): skip tool-declaration-only messages in non-Kimi providers
Message-level tool declarations (messages[].tools) are a Kimi wire
feature. The other providers' explicit field construction already keeps
the tools field off the wire, but the content-free leftover message
would be rejected (OpenAI: system message without content) or serialize
as a garbage <system></system> turn (Anthropic/Google system-to-user
wrapping). Skip such messages entirely via a shared predicate; a message
that also carries content only loses the tools field, as before.
Unreachable in kimi-code (the projection gate strips dynamic-tool
context for models without the select_tools capability before any
provider sees it) — defense-in-depth for direct kosong consumers.
* fix(agent-core): survive runtime flag flips and align tool table with post-compaction state
Two fixes from PR review:
- Register select_tools unconditionally and gate only its exposure in
loopTools. The tool-select flag can flip at runtime (config reload
calls setConfigOverrides on the live resolver) without
initializeBuiltinTools re-running; previously the disclosure shape
activated while the tool itself was unregistered, cutting the session
off from MCP entirely until a model/cwd change rebuilt the builtins.
A profile listing the name explicitly still never surfaces it in
inline mode, and execution guards the flip race defensively.
- Resolve the per-step tool table AFTER beforeStep, next to
buildMessages. beforeStep can run full compaction, which trims loaded
schemas and rewrites the ledger; a table captured before it could
still dispatch a tool whose schema the model no longer has. The
executable table and the request messages now always reflect the same
state, so a trimmed tool is rejected with select guidance instead of
executed.
* fix(agent-core): drop unused Tool import in dynamic-tools
* fix(agent-core): baseline compaction guard after post-compaction reinjection
The reinjected reminders (loadable-tools manifest, goal) are re-appended
after every compaction, but the nothing-new-since-compaction baseline was
captured before injectAfterCompaction. With a large manifest the guard
could re-trigger auto-compaction against a floor that cannot shrink.
Raise the baseline to the true post-compaction floor once reinjection
completes; the earlier capture stays as a fallback when reinjection
throws.
---------
Co-authored-by: fengchenchen <fengchenchen@moonshot.ai>
Default `thinking.keep` to "all" when Thinking is on so prior `reasoning_content` is kept across turns. Add `[thinking] keep` to config.toml and keep `KIMI_MODEL_THINKING_KEEP` as an override (env > config > default); off-values disable it.
* fix(tui): keep input anchored after slash command menu closes
Force a full re-render when the slash command menu closes, but only when the session content already overflows one screen; skipped under tmux. Detect the close edge from a render frame so asynchronous closes (Backspace deleting the leading slash) are covered too. Apply the same overflow/tmux gating when restoring the editor from selector panels.
* fix(tui): measure overflow against restored editor tree
Address Codex review: the overflow probe ran before the editor container was swapped back, so it counted the tall replacement panel and forced a full clear/home even when the restored content fit on one screen, yanking the editor to the top. Measure after the editor is mounted instead.
* fix(tui): redraw when content exactly fills one screen
Address Codex review: the guard skipped the forced redraw when the post-close layout is exactly one screen, leaving the editor shifted up because the differential renderer keeps the old viewport offset after a shrink. An exact fill is safe to clear (no blank tail), so redraw when content fills or overflows the viewport, and cover it with a test.
This package only ever contained a package.json with no sources, dependencies, or scripts, and nothing in the repo imports it (the CLI uses @moonshot-ai/migration-legacy instead). Remove the directory and drop its entries from flake.nix and the changeset config, then refresh the lockfile.
* fix(tui): complete @ file mentions across additional workspace roots with fd
When additional workspace directories are added via /add-dir, @ file completion fell back to a readdir-based scanner capped at 2000 entries, so deeply nested files in large projects never appeared. Route @ completion through fd across every root instead, keeping the query pushed down to fd and deduplicating by absolute path. The readdir fallback remains for when fd is unavailable.
* fix(tui): preserve per-root full-path fallback for @ mentions
Address review feedback: decide the scoped-versus-full-path fallback per root instead of globally. When one root has the scoped directory but another does not, the latter still runs a whole-tree --full-path search with the original query, so a match that only exists under that root is not hidden just because a sibling root happens to contain the prefix directory.
* fix(tui): fall back to filesystem when fd binary is not executable
Address review: when fdPath is non-null but the binary cannot be spawned (managed fd removed or lost execute permission), @ completion returned null because pi-tui swallows the spawn error into an empty result, so the catch never ran. Probe fd with accessSync(X_OK) before delegating and use the filesystem fallback when it is not executable, while still returning null for genuine no-match results.
* fix(tui): trust bare fd command names when probing executability
Address review: when fd is discovered on the system PATH, detectSystemFdPath returns the bare name (fd/fdfind). accessSync checked that literal string relative to cwd and never searched PATH, so a valid system fd was treated as unavailable and @ completion fell back to the capped scanner. Trust bare names (spawn resolves them via PATH) and only probe absolute/relative paths, which is how the managed fd is referenced and which can go stale.
* chore: remove accidentally committed plan files
* test(pi-tui): stabilize paste-burst test by freezing the clock
The paste-burst heuristic uses an 8ms inter-character interval that a slow or busy CI runner can exceed between synchronous handleInput calls, which resets the burst and lets Enter submit. Freeze Date so the synchronous keystrokes always register as one burst, making the assertion deterministic.
* chore: ignore top-level plan directory
A top-level Teleport in the sidebar template made the component multi-root, so v-show could not apply display:none and the collapsed sidebar stayed mounted at the rail width, squeezing the conversation. Move the teleport inside the aside so the sidebar is single-root again.
* feat(web): render AskUserQuestion result as an option list
Parse the AskUserQuestion tool result and echo the full option list, highlighting the chosen option(s) and dimming the rest, instead of dumping raw JSON. Handles single/multi select, free-text Other answers, and the dismissed state. Answers are zipped back to the input questions by index since the input carries no ids.
* fix(web): drop the stray indent in the tool-call card body
The expanded body of every tool-call card carried a hard-coded 36px left indent (a magic number meant to align with the header name). Drop it so expanded content aligns with the header's 11px padding instead. The design-system preview page mirrored the same rule and is updated to match.
* fix(web): align markdown diff block with the design system
Restyle the local ```diff renderer in chat Markdown to match the ~/diff panel: code text keeps the normal ink colour, the +/- sign carries the add/del colour, and rows use a soft background with an inset accent bar instead of dyeing the text green/red. The header and copy button now match the standard code-block chrome (height, hover, focus ring).
* fix(web): match the markdown diff block chrome to code blocks
Make the local markdown diff renderer's shell identical to a regular code block: swap the text copy button for the same icon button, inherit the header's text-xs sizing, and keep the container / header / code-area padding, background, radius and shadow in lockstep with .code-block-container and .code-block-header. The diff-specific row tinting (soft background, inset accent bar, coloured sign) is kept since that is what makes it a diff.
* fix(web): fall back to raw output for non-answer AskUserQuestion results
Background launches and error cases return plain-text tool output (task_id/status lines, or a failure reason), not the { answers } JSON the card expects. The card used to render an empty, fully-unselected option list in those cases, hiding the task id or error. Now the card only renders the option list when the output parses as the answer payload; otherwise it shows the raw output. Addresses review feedback (P2).
* fix(web): localize AskUserQuestion result labels
The card hard-coded user-visible strings (Dismissed, answer/answers, Answered, the (+N more) summary), so Chinese-locale transcripts mixed in English. Move them into the en/zh tools locale files and read them via t(). Addresses review feedback (P2).
- add GET /api/v1/workspaces/{workspace_id}/skills backed by a listWorkspaceSkills core RPC and ISkillService.listForWorkDir, reusing the session skill-loading path so results match a new session
- web: populate the composer slash menu from workspace skills before a session exists, then fall back to session skills once one is active
* feat(session): rebuild session index on boot and self-describe workDir
- persist workDir into state.json so session dirs are self-describing and
summaries do not depend on the index's one-way-hashed workDir
- relax readSessionIndex so a stale or non-absolute index workDir no longer
drops an otherwise valid entry
- serialize in-process index appends to avoid torn jsonl lines
- add SessionStore.reindex() and run it once at server boot so the
scan-free request path can find sessions whose index line is missing
or stale
* chore: add changeset for session index rebuild
* fix(session): repair index entries with a stale workDir during reindex
- send tool declarations, system prompt, and sampling/thinking settings in the camelCase shape the Google SDK forwards so tool calls reach the model
- thread tool-call extras (thought signatures) through the loop tool-call event into context so Gemini 3 can resume a tool turn
- update and add tests for the corrected request shape and signature round-trip
* fix(web): refill attachments when editing a queued or undone message
Queued prompts that carry images/video are no longer remove-only: clicking one loads its text and attachments back into the composer. Undo ("edit & resend") now restores the message's attachments too, not just its text. useAttachmentUpload gains loadAttachments, which reuses the existing fileIds (no re-upload) and fetches authenticated blob URLs for protected getFileUrl previews so the refilled thumbnails don't 401.
* fix(web): forward attachment refills through ChatDock
When the normal chat dock is mounted (non-empty conversation), bindChatDock receives the ChatDock instance, but ChatDock only exposed loadForEdit/focus — so the new loadAttachmentsForEdit fell back to a no-op and editing a queued media prompt or undoing a media message still dropped the attachments. ChatDock now forwards loadAttachmentsForEdit to the underlying Composer.
* fix(web): replace composer attachments when refilling edits
loadForEdit(text) overwrites the composer text, but loadAttachments appended to any unsent draft attachments, so a later submit sent the stale draft files together with the edited message's files. Make loadAttachments replace the current session's attachments (revoking their object URLs) so an edit/undo replaces the whole composer, mirroring loadForEdit.
* fix(web): clear stale attachments on text-only edits
When the composer already has attachments loaded (for example after editing a queued media prompt) and the user then edits a text-only queued prompt or undoes a text-only message, loadComposerForEdit skipped loadAttachmentsForEdit (the only path that clears the strip) because the new attachment list was empty/undefined, so the next submit would send the stale media with the new text. Always call loadAttachmentsForEdit(attachments ?? []) so text-only edits replace the strip with an empty set.
* fix(web): make fileId-less refilled media resendable
When editing a user turn whose media was base64-inlined by the server (no fileId), loadAttachments used to add a non-uploading chip with no fileId, which handleSubmit silently drops on resend (and an image-only edit would not submit at all). Re-upload the data URL to obtain a fileId so the attachment is actually resendable; when re-upload is unavailable, skip the chip instead of showing a misleading ready attachment. Also factor a patchAttachment helper.
* chore: prefix web changeset entry
The gen-changesets rules require web-app changelog entries to start with 'web: ' so the synced release notes classify web UI changes correctly.
* fix(web): don't dequeue a prompt when the composer is hidden
When a queued media item is clicked while the dock is showing a pending question or approval, ChatDock has no nested Composer (only rendered in its v-else), so loadComposerForEdit no-ops — but handleEditQueued still dequeued the item, losing it. Make ChatDock.loadForEdit report whether the nested composer is present, have loadComposerForEdit return success, and only dequeue when the load actually succeeds.
* fix(web): preserve URL-backed media when refilling composer
When an undone turn contains media with source.kind === 'url' (a URL but no fileId), loadAttachments used to fall through and drop it. Re-upload the URL (data: or http(s):) to obtain a fileId so URL-backed media is preserved on edit/resend; if the URL can't be fetched (CORS / non-2xx) the chip is dropped instead of shown as a misleading ready attachment.
Background (detached) shell commands were exempt from the 16 MiB output
ceiling, so a runaway background command could fill the disk or crash
the process. Apply the same cap to background shell commands and stop
feeding the disk write chain once it trips. Scope the ceiling to
process tasks so subagent and user-question results, which are appended
once and must be persisted, are left untouched.
* feat: hold print-mode turn until background subagents drain
In `kimi -p` (print mode), when the main agent ends a turn while background
subagents (`kind === 'agent'`) are still running, hold the turn open and
idle-wait until they finish, flushing their completions into the turn so the
model can react before the run exits.
Previously, the main agent could end its turn after launching background
subagents; the print flow then drained them with their completion
notifications suppressed, so the main agent never saw the results and the run
exited with the work abandoned (e.g. no nomination). This was the root cause
of the swarm-alpha-mining eval failures.
The hold is gated on a new `drainAgentTasksOnStop` session option (set by the
print flow), only affects `kind === 'agent'` background tasks, and is bounded
by `background.printWaitCeilingS`. Backfill / fan-out is handled by
re-enumerating active tasks. Other background task kinds and non-print modes
are unaffected.
Add the two kimi server run flags introduced in #1368 to the CLI reference (en + zh), including a danger callout for the auth-bypass flag, and add the changeset so the next release notes the feature.
* feat(server): add --dangerous-bypass-auth and --keep-alive flags
- --dangerous-bypass-auth disables bearer-token auth on every REST and
WebSocket route and advertises it via /api/v1/meta so the web UI skips
the token prompt; the startup banner drops the token and shows a red
danger notice
- --keep-alive keeps the daemon running instead of idle-killing after 60s;
implied by --host / --allowed-host and always on in --foreground mode
* fix(server): address review feedback on bypass-auth
- keep the token and skip the bypass notice when a daemon is reused, since
the requested --dangerous-bypass-auth flag is not applied to the
already-running server
- clear the cached dangerous_bypass_auth web state on HTTP 401 so a stale
bypass value cannot hide the token prompt after the server restarts
without the flag
* fix(pi-tui): make the viewport anchor follow above-viewport content shifts
The anchor pins a buffer row index, but an above-viewport length change
shifts the content living at every index below it. The pinned window
then suddenly showed content further along (visible upward creep), and
the rows that slid above the window top were lost: never committed to
scrollback, which holds older bytes at those indices. During streaming,
every above-viewport net shrink (a finished agent row collapsing, a
merged step, a spinner line disappearing) permanently swallowed that
many rows, and the blank area under the input box kept growing.
doRender now scores two hypotheses for such frames — window stayed put
vs window content shifted by the length delta — and when the shift
explains the frame strictly better, moves the anchor with the content:
a pure shift re-anchors with no painting at all (the screen already
shows exactly that content), and a shift with local in-window changes
(spinner/timer rows) repaints the window at the shifted anchor.
Commit order stays continuous, so the exactly-once scrollback invariant
is preserved: no loss, no duplication.
Covered by e2e case07 (above-viewport shift), which fails on the
previous revision; the stale-content unit test now asserts the
follow-the-content window instead of the old swallowed-row behavior.
* revert(pi-tui): restore upstream differential rendering behavior
The fork's viewport/scrollback rendering patches (clamping the diff to
the visible viewport, viewport re-anchoring on collapse, the pinned
anchor with commit-on-advance, cursor visibility guarding, and the
content-shift anchor follow) accumulated interacting edge cases faster
than they could be stabilized: blank screens, duplicated scrollback
spans, vanished rows, and a growing blank area under the input box.
Revert src/tui.ts to the upstream 0.80.2 differential rendering
behavior: a change above the viewport triggers a destructive full
redraw again. Verified line-by-line against the upstream source — the
only remaining divergences are the TypeScript strict-mode syntax
adaptations and the narrow-terminal fixes (Container width clamping
and overwide-line truncation replacing the upstream crash-and-throw),
which are kept.
Also remove the rendering-bug e2e ledger and the shrink test suite
that specified the reverted behavior, restore the pre-fork rendering
tests (the transient-content test asserts the upstream full-redraw
behavior again), and drop the e2e glob from the test script. Editor
input-history and paste-burst changes are untouched.
Known trade-off, accepted for now: the original scroll-position yank
during streaming (destructive redraws emitting ESC[3J) returns; the
rendering rework will restart from this clean baseline.
* chore: downgrade the web thinking-effort changeset to patch
* feat(web): support multi-level thinking effort selection
Surface each model's declared reasoning efforts (support_efforts /
default_effort) in the web model picker as a segmented control,
replacing the on/off toggle. ThinkingLevel is now an open string, the
model catalog carries the effort metadata to the web app, and the
mobile settings sheet and /thinking command cycle through the
available levels.
* fix(web): preserve persisted thinking level before models load
coerceThinkingForModel returned 'on' for any non-'off' level when the
active model was still undefined (catalog not loaded yet), which
rewrote a persisted/default effort like 'high' to 'on' and silently
dropped the model's declared effort on later prompts. Keep the
requested level as-is until loadModels() re-runs coercion with the
real model.
Addresses Codex P1 review on modelThinking.ts.
* fix(web): coerce stale thinking level against the active model
When selecting another session, the persisted thinking level can be a
boolean 'on'/'off' carried over from a previous model. Deriving the
pill suffix and active segment from that raw value could show
"thinking: off" on an always-on model or hide the concrete effort
behind a bare "thinking" tag. Coerce the level against the current
model before deriving display state.
Addresses Codex P2 review on Composer.vue.
* fix(web): submit coerced thinking level for the prompt's target model
The send paths (submitPromptInternal / steerPrompt) previously submitted
rawState.thinking verbatim, so a value carried over from another session
(e.g. 'max' from an effort model) was sent to a model that doesn't
declare it, even though the composer already showed the coerced default.
Coerce the level against the target model before submitting so the first
turn runs with the level the UI displays.
Addresses Codex P2 review on Composer.vue.
* fix(web): coerce stale thinking in /thinking and mobile settings
Two more surfaces derived their active state from the raw persisted
level, which is stale when the saved level came from a different model:
- /thinking slash command: indexing the raw value (e.g. 'on' from a
boolean model) into an effort model's segments returned -1 and jumped
to 'off' instead of advancing from the model's default effort.
- Mobile settings sheet: clamping the raw prop fell back to the first
segment, showing/selecting 'off' or the first effort while the
composer and prompt submission coerce to the model default.
Coerce the level against the active model in both places, matching the
composer and send-path behavior.
Addresses Codex P2 reviews on App.vue and MobileSettingsSheet.vue.
* feat(media): materialize video uploads to cache and reference by path
- copy TUI video placeholders into the shared cache instead of
inlining the original source path
- emit <video path="..."> tags so ReadMediaFile / the provider's
VideoUploader owns upload behavior
- apply the same cache materialization to server prompt video
submissions, matching the TUI flow
- update TUI unit tests and server e2e test to assert cache-path
behavior
* fix(web): make uploaded videos play in the chat
Render the server's <video path> tag as a real video and reconcile the echoed user message so the bubble no longer shows raw markup or a duplicate. Serve file downloads with byte-range support and fetch video bytes with the bearer credential into a blob URL, since browsers cannot authorize a <video> src on their own. Also let users click an uploaded image to open it in the preview panel.
* fix(web): use authenticated source for uploaded image previews
openMediaPreview stored the raw getFileUrl as sourceUrl, and FilePreview renders it with a native <img> that sends no Authorization header, so the enlarge action 401'd for uploaded images. When the media carries a fileId, fetch the bytes through the authenticated API client and preview a blob URL instead, revoking it when the preview is replaced or closed.
* fix(web): ignore stale authenticated media fetches
AuthMedia fetches the file bytes asynchronously; when the component is reused with a new fileId before a prior fetch resolves (e.g. queued thumbnails keyed by index), the older response could still create a blob URL and show the previous file. Add a per-request sequence guard (and an unmount guard) so a stale response is discarded and its blob URL revoked instead of being applied.
* fix(web): gate media path tags on file-store id shape
Treating any standalone <video path="..."> text as an uploaded daemon file and stripping the basename into getFileUrl is only valid for server cache files named after the file-store id (f_…). TUI/ReadMediaFile tags use arbitrary cache names like <uuid>-<label>, and older transcripts may point at paths like /tmp/foo.mp4; those produced a broken /files/<basename> request. Only extract a fileId when the basename matches the file-store id shape, otherwise leave the raw tag as text.
* fix(web): invalidate pending media preview on close
Closing an uploaded-image preview before getFileBlob() resolved left previewRequestSeq untouched, so the fetch callback still passed its seq check, created a blob URL, then skipped attaching it because previewFile was already null — leaking up to the file size until another preview opened. Bump previewRequestSeq on close so the in-flight callback bails before creating the blob URL.
* fix(web): defer authenticated media fetch until near viewport
AuthMedia fetched the full image/video into a Blob on mount whenever a fileId was present, bypassing native loading="lazy" and preload="metadata". Opening a session with several historical large video uploads started many full downloads and held all blobs in memory even if the user never scrolled to or played them. Use an IntersectionObserver to defer the fetch until the element nears the viewport.
* fix(web): revoke preview blob when leaving the file panel
Switching to another detail panel only flips detailTarget and never calls closeFilePreview, so an in-flight getFileBlob could still create a blob URL after the file panel hid, and an already-shown blob URL was held until the next file preview. Check detailTarget before creating the blob URL, and reset/revoke the preview when detailTarget leaves 'file'.
---------
Co-authored-by: haozhe.yang <yanghaozhe@moonshot.ai>
* fix(pi-tui): stop scrollback duplication from viewport rewinds
Rewinding the viewport anchor repaints rows that the terminal
scrollback already holds, and the next scroll commits them again —
every rewind duplicated its span. Two paths triggered it during
streaming oscillation (content shrinking then growing back): the
shrink re-anchor rewound immediately, and the clamped differential
path then painted the shifted content through the screen.
Rework the shrink/shift handling around a shared in-place viewport
repaint that never scrolls, with the anchor treated as the scrollback
high-water mark that must never move backward:
- Partial shrinks keep the anchor pinned: the content bottom hovers
above a bounded blank gap that the next growth naturally fills. No
rewind means duplication is impossible by construction.
- Only a collapse past the viewport top (compaction, clears) rewinds,
as nothing sensible could be shown otherwise; the content has
changed so drastically there that the repainted span is not
recognizable as a duplicate.
- Above-viewport length changes repaint the visible window in place
instead of painting through: nothing scrolls, scrollback keeps the
stale old version, and the anchor only advances when the content
outgrows the pinned window.
- Deleted-tail changes within a pinned viewport repaint at the pinned
anchor instead of falling back to a destructive full render.
Pure appends keep flowing through the screen into scrollback, and
equal-length above-viewport changes keep the bounded clamped diff.
* fix(pi-tui): commit skipped rows on anchor advance and guard cursor visibility
Address two review findings on the pinned-anchor rendering:
- An anchor advance (growth past the pinned viewport combined with an
above-viewport change) repainted the screen in place without
scrolling, so the rows between the old and new anchor were never
committed to scrollback and vanished. repaintViewport now paints from
the old anchor and lets the paint loop scroll the skipped rows out,
committing each exactly once with fresh content.
- positionHardwareCursor recorded hardwareCursorRow on a logical row
outside the visible window when the cursor marker sat above a pinned
viewport (tall editor after a deep shrink), desyncing every later
differential move. It now hides the cursor and keeps the bookkeeping
on the real cursor row when the marker is not visible.
Also add an e2e rendering-bug ledger (packages/pi-tui/e2e): one
xterm-emulated repro per production rendering bug, asserting the
renderer invariants (monotonic anchor, exactly-once commit, cursor
bookkeeping sync). The two findings above are case05/case06.
* ci: run the pi-tui node:test suite in a dedicated job
pi-tui's tests (unit + e2e) run on node:test, which the root vitest
run silently skips, so CI never executed them. Add a test-pi-tui job
that runs the package's test script.
* feat(agent-core): guide the model away from repeating denied or failed tool calls
- system.md: add a diagnose-before-retrying paragraph next to the existing
permission-denial guidance, covering failed tool calls
- permission: when the user rejects an approval on the main agent, tell the
model not to re-attempt the exact same call (sub agents already had an
equivalent hint)
* fix(agent-core): close abandoned tool exchanges and dedupe duplicate tool_use ids
A turn that dies between a recorded tool.call and its paired tool.result
(e.g. a transcript write failure mid-batch) used to leave
pendingToolResultIds open forever: every later message was stranded in
deferredMessages and user input was silently swallowed.
- runOneTurn now defensively closes any dangling tool calls when a turn
ends (completed, cancelled, or failed), synthesizing an error result
that names the cause, with a warn log and a tool_exchange_abandoned
telemetry event
- the projector drops assistant tool calls whose id already appeared
earlier (first occurrence wins): a duplicate id is wire-invalid on
strict providers and not repairable by the strict resend; reported via
the existing projection-repair log and telemetry
- resume-side closePendingToolResults now logs what it closes (warn for
a mid-history gap, info for the routine trailing interruption)
* chore: add changesets for tool exchange fixes
* fix(agent-core): scope duplicate tool_use id dedup to the strict resend
Unconditional dedup regressed providers that emit per-response counter
ids (e.g. call_0 in every step) and accept their own duplicates: later
tool exchanges silently vanished from the projected history, and a
duplicate call's own recorded result was left dangling.
- the dedupe pass is now opt-in via dedupeDuplicateToolCalls and enabled
only in strictMessages, so the normal projection keeps the history the
provider produced
- the pass also drops every tool result after the first for an id, so no
dangling tool message survives; when the kept call has no result of
its own, the surviving one is reattached by the adjacency repair
- kosong now classifies the Anthropic "tool_use ids must be unique" 400
as a recoverable request-structure error so it triggers the strict
resend
* feat(cli): wait for background subagents before exiting kimi -p
When `background.keep_alive_on_exit` is enabled, `kimi -p` now waits for
all background subagents to reach a terminal state before exiting, bounded
by `background.print_wait_ceiling_s` (default 3600s). This lets concurrent
background subagents run to completion in single-turn runs instead of being
torn down when the main agent's turn ends.
Stable height across running, done, failed and backgrounded states: all share the same header + one-line tool summary + two-row content window, so the card no longer shrinks when a run finishes. Add a braille spinner in the header while active, collapse sub-tool calls into a one-line summary, and have the two-row window follow the live stream (tool output, text, or thinking) instead of showing thinking and text side by side. Mute the window tones so a brief text or tool-output segment no longer flashes white against dim thinking.
* fix(agent-core): route image-compression captions through hidden system reminders
Prompt ingestion (server upload/base64 route, TUI paste, ACP) annotates a
compressed image with an inline <system> caption inside the user's own
message. That raw markup rendered verbatim in every user-visible history
projection (TUI session replay, web UI) and leaked into session titles.
Split the caption out at the appendUserMessage chokepoint and deliver it
through the built-in system-reminder injection (origin
{kind: 'injection', variant: 'image_compression'}), which every UI already
hides. The model still receives the full note; ingestion sites and the wire
protocol are unchanged. Session titles/lastPrompt strip the caption the same
way. Tool-result captions (MCP) keep the established <system> convention.
Covered by unit tests plus an end-to-end smoke suite that drives
rpc.prompt/steer through the real turn pipeline and asserts the provider
wire request, stored history, replay records, and resume parity.
* chore: tighten changeset wording per gen-changesets conventions
* feat(agent-core): strengthen the language-matching rule in the default system prompt
* chore: refine changeset wording
* fix(kaos): enrich PATH from the user's login shell at startup
When kimi-code is launched from a context that skipped the user's shell
profile (GUI launchers, non-login parent shells), process.env.PATH misses
entries like /opt/homebrew/bin, so commands spawned by the Bash tool
cannot find user-installed tools such as gh.
LocalKaos.create() now probes the user's login shell once
($SHELL -l -c env, 5s timeout, memoised) and appends the missing PATH
entries to process.env.PATH. Existing entries keep their order and
priority; probe failures silently leave PATH untouched. Windows is
skipped: the problem is specific to POSIX login-shell profiles.
* fix(kaos): fall back to the account login shell when $SHELL is unset
launchd/daemon launches can leave $SHELL unset or blank — the very
contexts whose PATH is impoverished — so the login-shell PATH probe
would give up exactly where it matters most. Resolve the shell from the
OS user database (os.userInfo().shell) before giving up; lookups that
throw (uid without a database entry) or yield nologin shells degrade
silently as before.
* fix(kaos): preserve empty PATH components when merging login-shell PATH
POSIX command lookup treats an empty PATH component (leading colon,
trailing colon, or double colon) as the current directory. The merge
previously filtered those out of the current PATH and rewrote the value
even when nothing was appended, silently dropping cwd lookup for users
who rely on it.
Keep the current PATH string verbatim as the prefix, append only the
missing login-shell entries, and skip the env write entirely when the
login shell contributes nothing — an unset PATH stays unset, a set PATH
is never rewritten. Empty login-shell components are still never
imported.
* fix(kaos): only import absolute login-shell PATH entries
A `.` or relative component in the login-shell PATH is cwd-dependent
lookup with another spelling, and LocalKaos runs commands from arbitrary
workspace directories — importing one would let a command name resolve
from an untrusted project cwd. Tighten the merge's skip condition from
"empty" to "not absolute", which subsumes the empty-component check.
* fix(kaos): invoke the login-shell probe's env by absolute path
A bare `env` inside `$SHELL -l -c` resolves through the inherited PATH
from the workspace cwd. If that PATH carries a cwd-dependent component
(which the merge deliberately preserves), a repo-planted `env` binary
would run automatically at session startup and could feed the probe an
arbitrary PATH. /usr/bin/env is guaranteed on mainstream POSIX systems
and also bypasses profile function shadowing.
* feat(agent-core): guide the model away from repeating denied or failed tool calls
- system.md: add a diagnose-before-retrying paragraph next to the existing
permission-denial guidance, covering failed tool calls
- permission: when the user rejects an approval on the main agent, tell the
model not to re-attempt the exact same call (sub agents already had an
equivalent hint)
* fix(agent-core): close abandoned tool exchanges and dedupe duplicate tool_use ids
A turn that dies between a recorded tool.call and its paired tool.result
(e.g. a transcript write failure mid-batch) used to leave
pendingToolResultIds open forever: every later message was stranded in
deferredMessages and user input was silently swallowed.
- runOneTurn now defensively closes any dangling tool calls when a turn
ends (completed, cancelled, or failed), synthesizing an error result
that names the cause, with a warn log and a tool_exchange_abandoned
telemetry event
- the projector drops assistant tool calls whose id already appeared
earlier (first occurrence wins): a duplicate id is wire-invalid on
strict providers and not repairable by the strict resend; reported via
the existing projection-repair log and telemetry
- resume-side closePendingToolResults now logs what it closes (warn for
a mid-history gap, info for the routine trailing interruption)
* chore: add changesets for tool exchange fixes
* fix(agent-core): scope duplicate tool_use id dedup to the strict resend
Unconditional dedup regressed providers that emit per-response counter
ids (e.g. call_0 in every step) and accept their own duplicates: later
tool exchanges silently vanished from the projected history, and a
duplicate call's own recorded result was left dangling.
- the dedupe pass is now opt-in via dedupeDuplicateToolCalls and enabled
only in strictMessages, so the normal projection keeps the history the
provider produced
- the pass also drops every tool result after the first for an id, so no
dangling tool message survives; when the kept call has no result of
its own, the surviving one is reattached by the adjacency repair
- kosong now classifies the Anthropic "tool_use ids must be unique" 400
as a recoverable request-structure error so it triggers the strict
resend
