Docker containers spawned by MCP servers configured with `docker run`
were not cleaned up after being stopped, causing exited containers to
accumulate and fill up Docker storage over time.
Closes#8103
Users launching a follow-up implementation session now get an explicit plan-file path and instruction to read it first, so execution stays aligned with the canonical plan even when copied plan text drifts.
Update hono catalog version from 4.10.7 to 4.12.12 to fix 14 advisories
including JWT algorithm confusion (GHSA-f67f-6cw9-8mq4, GHSA-3vhc-576x-3qv4),
CORS bypass, body limit bypass, XSS, cookie injection, SSE injection,
path traversal, and prototype pollution.
Add null guard for ptyID param in pty.ts to satisfy hono 4.12's stricter
return type for c.req.param().
* refactor(vscode): enrich migration session data with title, directory, and time metadata
Replace plain session ID strings with structured MigrationSessionInfo objects
throughout the legacy migration pipeline. Session detection now extracts
task title, workspace directory, and timestamp from global state, providing
richer context for the migration wizard while still passing only IDs when
executing the actual migration.
* feat(vscode): add granular session progress tracking to legacy migration pipeline
Introduce MigrationSessionProgress type and SessionProgressCallback to
emit per-session phase updates during migration. The migration handler
now forwards structured progress events—including session metadata,
index, total count, and current phase—to the webview via a new
legacyMigrationSessionProgress message type, enabling the UI to display
detailed real-time feedback for each session being migrated.
* refactor(legacy-migration): extract session progress helpers and wire per-phase reporting into migrate
Move inline progress construction out of the migration loop into
dedicated buildSessionMeta and buildSessionProgress factory functions
in a new migration-session-progress module. Update the migrate function
signature to accept optional meta and progress callback, then emit
fine-grained phase events (project, session, messages, parts, skipped,
done, error) with current/count counters so callers receive incremental
feedback as each import step completes.
* feat(vscode): display per-session migration progress in MigrationWizard
Add SessionProgressState interface and wire up handling for the
legacyMigrationSessionProgress message type so the wizard UI can
render real-time phase, index, and error details for each session
being imported during legacy migration.
* refactor(migration): extract SessionMigrationProgress into dedicated component
Move SessionProgressState interface and progress rendering logic out of
MigrationWizard into a new SessionMigrationProgress component with its
own state type, step pipeline visualization, and associated CSS styles.
- Add SessionMigrationProgress.tsx with step-dot UI for project,
session, messages, and parts phases
- Replace inline SessionProgressState with exported
SessionMigrationProgressState from the new module
- Add stepper styles to migration.css with active/success dot states
- Re-export SessionMigrationProgress from the migration barrel
* fix(vscode): improve session migration progress UI and timing
Add inter-session delay to prevent rapid-fire migration requests,
fix session spinner icon to remain active while sessions are still
being processed, and restructure the progress metadata layout.
- Add 300ms delay between session migrations in migration-service
- Keep sessions group spinner active until all sessions complete
- Split meta line into separate directory, title, and date elements
with proper truncation and ellipsis handling
- Reformat date display to show time before date
- Show "Session skipped" label when phase is skipped
- Tighten spacing and padding in session progress card
* feat(migration): add post-migration summary view for session progress
Introduce a summary phase that displays aggregated migration results
after all sessions have been processed. The summary card shows counts
of imported, skipped, and errored sessions along with the last error.
- Add "summary" phase to MigrationSessionPhase union type
- Emit summary progress event with 1s delay after final session
- Create SessionMigrationCard wrapper component for shared card styling
- Create SessionMigrationSummary component with imported/skipped/errored
breakdown
- Add session-migration-summary-state module with pure reducer logic
for accumulating per-session outcomes
- Extract card border/background styles into reusable .migration-session-card
- Switch progress view to summary view when phase reaches "summary"
* refactor(legacy-migration): defer finalization to explicit client message
Move post-migration side effects (global dispose, status update,
broadcast, session refresh) out of handleStartLegacyMigration and
into a new handleFinalizeLegacyMigration handler. The webview now
sends a "finalizeLegacyMigration" message when the user clicks Done,
decoupling migration execution from completion bookkeeping.
- Extract finalization logic into handleFinalizeLegacyMigration
- Add FinalizeLegacyMigrationMessage type and wire into WebviewMessage union
- Register "finalizeLegacyMigration" case in KiloProvider message switch
- Trigger finalization from MigrationWizard handleDone before clearing data
* fix(migration): move error copy action into session summary and preserve raw error text
Remove the standalone error box and clipboard helper from
MigrationWizard, relocating the copy-to-clipboard functionality
directly into SessionMigrationSummary next to the LastError label.
Store the full error string as lastErrorRaw on SessionSummaryState
and display only the first line as the visible lastError.
- Add lastErrorRaw field to SessionSummaryState and short() helper
- Inline copy button and toast into SessionMigrationSummary component
- Add migration-session-summary__row layout style for label + button
- Remove redundant error-box markup and copySessionError from wizard
* refactor(legacy-migration): expand summary view with per-session detail lines and structured error formatting
Replace aggregate session counts with individual item listings
using a scrollable list for each category (imported, skipped,
errored). Extract formatting logic into a dedicated module
(session-migration-summary-format.ts) that handles date
formatting, single-line display, and full error report
generation for clipboard export.
- Add session-migration-summary-format.ts with line, short,
errors, and copy helpers
- Render each session item via `<For>` instead of showing counts
- Introduce scrollable __list and truncated __item CSS styles
- Change copy button to aggregate all errors rather than a
single lastErrorRaw value
* feat(migration): add full summary report export with detailed error blocks
Introduce a `report` function that compiles all migration categories
(successful, skipped, errored) into a single clipboard-exportable
document. Each error entry now includes session metadata (id,
directory, title, date) alongside the raw error message.
- Add `report()` to session-migration-summary-format.ts assembling
a complete summary with per-category blocks
- Enrich `errors()` output with structured session metadata fields
- Relocate copy button to summary header and change label to
"Copy Report" to reflect full-report semantics
- Remove error-section-scoped copy button in favor of top-level action
* style(migration): improve summary list visuals with bordered sections and empty state placeholders
Add color-coded borders per category (success, skipped, errored),
render "None" placeholders when lists are empty, apply subtle
background styling to list containers and items, and truncate
long directory paths to their last segment for readability.
* refactor(migration): inline error details beneath errored sessions and simplify report format
Replace the separate "Errors" section with per-session error detail
lines displayed directly under each errored entry. Introduce a
`detail` helper that extracts the first line of an error, render it
in a new `__entry`/`__detail` layout with truncation styling, and
consolidate the export report to embed short error text inline
rather than appending a standalone error block.
* style(migration): restyle summary lists as flat striped rows with dynamic section labels
Flatten list items by removing per-item backgrounds and border-radius
in favor of alternating zebra-stripe rows with subtle separators.
Extract an `errored` helper to produce a flat row/detail stream and
add a `label` function that appends counts and optional descriptions
to section headings. Adjust font sizes, padding, and gap values for
a denser, table-like appearance.
* feat(migration): add force-reimport support for session migration
Introduce a `force` flag on session selections that allows previously
migrated sessions to be re-imported by deleting the existing row and
reinserting with updated data. Change session selections from plain
string IDs to `MigrationSessionSelection` objects carrying `id` and
an optional `force` field. Update the backend import service to use
`onConflictDoUpdate` instead of `onConflictDoNothing`, ensuring
forced reimports fully overwrite stale session records.
* fix(ui): add selectable skipped-session list with force re-import action
Introduce checkbox selection UI for skipped sessions in the migration
summary, allowing users to pick individual or all skipped sessions and
trigger a force re-import. Wire selected IDs through `postMessage` to
the backend with the `force` flag enabled. Add custom checkbox marks,
a "select all" toggle, and supporting CSS for the new interactive
controls.
* feat(migration): introduce preparing phase and improve error handling in session migrate
Add a "preparing" step to the session migration pipeline so users see
feedback while the session is being parsed. Move `parseSession` inside
the try block and re-parse on failure so the payload is always available
for error reporting. Extract a shared `fail` helper to eliminate
duplicated error-handling logic. Update phase types, step definitions,
and progress UI across backend and webview to reflect the new phase.
* refactor(migration): simplify session migration phases and remove granular progress tracking
Collapse the project, session, messages, and parts phases into a single
"storing" phase and drop the current/count progress fields. The
fine-grained per-message and per-part counters added unnecessary
complexity without meaningful UX benefit. The pipeline now reports only
two steps—preparing and storing—making the progress UI cleaner and the
data contract between backend and webview smaller.
* fix(migration): skip already-imported sessions by checking existence before parsing
Add a pre-flight check that queries for an existing session using the
legacy ID before performing any parsing or storing work. When the session
already exists and the force flag is not set, the migration short-circuits
to a "skipped" result immediately. This avoids redundant parsing and
duplicate writes for sessions that were previously imported.
* feat(ui): add confirmation dialog before force re-importing skipped sessions
Introduce ForceReimportDialog component that warns users about
overwriting previously migrated sessions. The SessionMigrationSummary
now opens this dialog on force re-import instead of immediately
dispatching the migration message, preventing accidental data loss.
* fix(migration): correct preparing step state resolution in progress tracker
Add missing condition to return "active" state when the preparing step
matches the current preparing phase, fixing the step indicator not
highlighting correctly during the preparation stage.
* refactor(ui): lift force-reimport logic into MigrationWizard and add running state guard
Move dialog orchestration and migration message dispatch for force
re-imports from SessionMigrationSummary up to MigrationWizard, keeping
the summary component presentational via a new `onForce` callback prop.
Introduce a `running` signal to track active migrations and add a
RunningMigrationDialog that warns users before leaving mid-migration.
Extend the `MigrationResultItem` category union with "session" so
re-import results can be filtered correctly.
* fix(migration): deduplicate session entries across all categories on state update
Extract a `strip` helper that removes a session from imported, skipped,
and errored lists before re-inserting it into the target category. This
prevents the same session from appearing in multiple summary buckets
when its status changes (e.g. a re-imported session that was previously
skipped or errored) and removes the early-return guard that silently
dropped valid "done" updates for previously-seen entries.
* style(migration): increase font size and spacing in session summary list
Bump item font size from 12px to 13px, detail font size from 11px to
12px, and adjust line-height, padding, and max-height values across
summary items, details, and interactive rows for improved readability.
* refactor(migration): reorder skipped-session actions and align layout to end
Move "Force Re-import" button after the "Re-import all" checkbox in
the actions bar, switch container justification to flex-end, and
remove min-width constraint from the select-all label for a tighter
trailing alignment.
* fix(migration): hide skipped and errored sections when their lists are empty
Wrap the skipped-session and errored-session sections in `<Show>` guards
so they are only rendered when items exist, removing the "None" fallback
placeholder and eliminating empty sections from the summary view.
* test(migration): add skip-detection and force-reimport coverage for session import
Introduce three new test files covering the duplicate-detection and
force-reimport paths added to the migration pipeline:
- migrate.test.ts: verify sessions are skipped before parsing when they
already exist, and that force=true bypasses the check; update call
sites to use the new { id, force } input shape
- session-summary-state.test.ts: validate that updateSessionSummary
correctly deduplicates entries across categories on state transitions
- session-import-service.test.ts: confirm the backend service returns
skipped when a session exists without force, and deletes then
recreates when force is true
* feat(i18n): internationalize migration dialog and summary strings
Replace hardcoded English text in ForceReimportDialog,
RunningMigrationDialog, SessionMigrationSummary, and MigrationWizard
with translation keys via useLanguage().t(). Add corresponding entries
across all 19 locale files covering session summary, force-reimport,
and running-migration labels.
* fix(migration): make error payload optional and remove redundant async
Make the `payload` field optional on failure results so errors thrown
before session parsing can be reported without a payload. Remove
unnecessary `async` from the `skip` helper and conditionally spread
`payload` only when present. In the outer catch block, call `fail`
without parsing the session since the parse itself may have been the
source of the error.
* refactor(migration): remove premature success toast from force reimport
The toast notification was firing immediately when the user triggered
a force reimport, before the operation actually completed. Remove it
from the click handler so feedback can be provided by the caller once
the reimport results are known.
* fix(migration): use live session list instead of stale cached reference
Replace `cachedSessions` fallback with the canonical `sessions` array
when resolving session metadata during migration. The cached reference
could hold outdated entries after a force reimport, leading to lookups
against stale data.
* refactor(migration): remove unused error-tracking fields and dead code
Drop `lastError`/`lastErrorRaw` from `SessionSummaryState`, the
`MigrationFinalization` interface, the `short()` helper, and the
`errors()`/`copy()` summary formatters. These were introduced for
inline error display but are superseded by the full `report()`
output. Update tests to reflect the simplified state shape and add
a clarifying comment to the session-import upsert logic.
* feat(i18n): add translation keys for migration progress and summary labels
Extract shared formatting helpers into `session-migration-format.ts`
and wire all hardcoded English strings in `SessionMigrationProgress`
and `SessionMigrationSummary` through the i18n layer. New keys cover
session progress steps, the migration counter header, summary section
labels, and fallback placeholders across all 20 supported locales.
* fix(migration): correct sessions group status to reflect active progress
Override the default status resolution for the "sessions" group so it
reports "migrating" while the migration is still running and individual
entries remain pending or in-progress, preventing the group from
prematurely settling to a terminal state.
* fix: formatting
* feat(migration): track error state and finalize with completed_with_errors status
Introduce `lastMigrationHadErrors` flag on MigrationContext to capture
whether any migration entry failed or the entire process threw an
exception. Use this flag during finalization to persist a distinct
"completed_with_errors" status instead of unconditionally marking the
migration as "completed", allowing downstream UI and logic to
differentiate between clean and partial migrations.
Health check endpoints generate high-volume, low-value log entries.
Add /global/health to the skipLogging set alongside /log and
/telemetry/capture to reduce log noise.
Pass the selected organization context through to KiloClaw API
endpoints so the /kiloclaw command resolves the correct org-scoped
instance when a team is selected via /teams.
* fix(cli): guard against null theme in TUI resolveTheme and Proxy
Add null guards to prevent crashes when theme.theme is null/undefined:
- Add isValidTheme type guard with fallback to built-in kilo theme
- Wrap theme resolution memo in try/catch for resilience
- Use empty object as Proxy target instead of memo result
- Validate custom theme JSON and protect built-in theme keys
Closes#8109
* fix(cli): validate required theme keys in isValidTheme
Check for background, text, and primary keys to reject empty or
partial theme objects that would produce undefined field accesses.
* core: fix revert diff showing incorrect changes after file restore
When reverting a session, the diff was computed after files were already
reverted to their pre-change state, producing an empty or wrong diff.
By computing the diff before restoring files, the revert summary now
correctly reflects what was actually undone for the user.
* core: reduce revert diff size in DB to save storage
Full file contents were being stored in the session summary diffs,
causing unnecessary database bloat. The webview RevertBanner only
needs file path, line counts, and status — not the full diff content.
* chore(cli): add kilocode_change markers to revert diff changes
---------
Co-authored-by: Marius <marius@kilocode.ai>
* fix(cli): prevent unbounded log file growth with size-based rotation
Fix two issues causing log files to grow to 50+ GB:
1. Server middleware leaked a "started" log line for skipped endpoints
because log.time() was called unconditionally — only timer.stop()
was gated by the skipLogging flag. Restructure to early-return for
skipped paths so no logging occurs at all.
2. Log files had no size limit within a session. A long-running
kilo serve process (e.g. VS Code extension) would write a single
file indefinitely. Add rotating-file-stream (50 MB, maxFiles: 1)
to cap log file size automatically.
* fix(cli): address review feedback on log rotation
- Revert server.ts changes (handled separately in #8141)
- Add error/warning event handlers on rotating-file-stream
- Increase maxFiles from 1 to 3 for better debug headroom
- Set explicit history file path to avoid .txt side-effect
* fix(cli): revert maxFiles back to 1
maxFiles: 1 is sufficient — we only need to cap disk usage, not retain
old rotated fragments. The existing cleanup() handles session-level
file retention separately.
* fix(cli): set maxFiles to 10 for log rotation
The MCP importer was silently ignoring disabled MCP servers during import.
Now imports all configured MCPs and preserves their disabled state by
setting enabled: false in the converted config.
Closes#8096
Co-authored-by: kiloconnect[bot] <240665456+kiloconnect[bot]@users.noreply.github.com>
Co-authored-by: Imanol Maiztegui <imanol.mzd@gmail.com>
Co-authored-by: Mark IJbema <mark@kilocode.ai>
* fix(cli): inject plan file path into non-experimental plan mode prompt
The plan.txt prompt said 'STRICTLY FORBIDDEN: ANY file edits, ZERO
exceptions' without telling the agent where the plan file is. The agent
had to guess from the directory tree, often picking .kilocode/ (which
exists) over .kilo/ (which doesn't). Now the non-experimental path
computes Session.plan(), creates the directory, and injects the exact
path into the prompt — matching what the experimental plan mode already
does.
Fixes#6907
* fix(cli): use absolute plan path in prompt to match experimental mode
Addresses bot review: relative path resolves against Instance.directory,
not Instance.worktree, so subdirectory sessions would target the wrong
file. Use the absolute path from Session.plan() like the experimental
plan mode does.
* chore(cli): add kilocode_change markers to plan file path injection
* fix: keep plan follow-up sessions in the same worktree
Route follow-up question replies by the owning session and create child implementation sessions from the planning session directory so "Start new session" stays visible and runs in the right worktree.
Refs #6339
* fix: keep worktree follow-up sessions reloadable
Adopt backend-created worktree sessions by normalized directory matching and keep plan follow-up sessions as roots so Agent Manager still restores them after reload.\n\nRefs #6339
* fix(vscode): track plan follow-up sessions in the sidebar
Record pending "Start new session" replies in single-session providers so the follow-up session.created event can be adopted and opened in the sidebar and regular tabs, while Agent Manager keeps using its own worktree adoption path.\n\nRefs #6339
* fix(vscode): forward pending follow-up session events
Allow pending plan follow-up session.created events through the KiloProvider SSE filter so sidebar and tab providers can adopt and open them before the new session is tracked.\n\nRefs #6339
* fix(agent-manager): open tab for adopted follow-up sessions
Navigate to the correct worktree or add a local tab when Agent Manager adopts a backend-created follow-up session, matching the existing sessionForked behavior.\n\nRefs #6339
* refactor(vscode): unify follow-up session adoption across all providers
Remove the separate adopt-session mechanism for Agent Manager and use the same KiloProvider follow-up path that sidebar and tabs use. The sessionCreated webview handler now works regardless of the current sidebar selection, so follow-up sessions appear as local tabs everywhere.\n\nRefs #6339
* test: align followup test with parentID field from main
The server middleware logs 3 lines per HTTP request (request, started,
completed). For /telemetry/capture this creates unnecessary log volume
since autocomplete telemetry can fire multiple events per keystroke
(Suggestion Requested, LLM Suggestion Returned, Unique Suggestion Shown).
Add /telemetry/capture to the skipLogging check alongside /log to reduce
noise in the kilo server log files at ~/.local/share/kilo/log/.
When the VS Code extension host is hard-killed (crash, OOM, kill -9),
dispose() is never called and the kilo serve child process is orphaned.
Fix by using a stdin keepalive pipe: the extension keeps stdin open after
spawning the CLI. The CLI watches for EOF on stdin (only when
KILO_CLIENT=vscode) and triggers a clean shutdown. The OS closes the
write-end of the pipe on any parent death, so the CLI always exits
regardless of how the extension host terminates.
On clean dispose(), stdin is explicitly closed before sending SIGTERM,
so the shutdown path is consistent in both the crash and clean cases.
Closes#8050
* feat(vscode): detect sessions in old extension
* feat(vscode): added session types
* feat(cli): add endpoints for inserting a session
* feat(vscode): use session types from sdk for parsing
* feat(kilo-vscode): parsing session project
* feat(vscode): parse legacy session details
* feat(kilo-vscode): refactoring
* feat(vscode): parse legacy session messages
* feat(vscode): parse legacy session parts
* feat(vscode): added testing files and id creation
* feat(vscode): merge tool parts on imported session
* feat(vscode): session message part refactoring
* feat(vscode): cover provider-specific reasoning parts on session migration
* feat(kilo-vscode): handle completion result parts
* feat: type hardening on session migration
* feat(kilo-vscode): create tool part correctly
* feat(kilo-vscode): adjust session part processing
* feat(vscode): persist migrated session in kilo.db
* feat(vscode): add error handling to session migration
* feat(vscode): add directory field to session migration
* feat(kilo-vscode): display completion check on session migration
* feat(vscode): prevent migrating sessions twice
* feat(cli): use correct project id when inserting migrated session
* feat(kilo-vscode): Use correct project id for session migration
* feat(kilo-vscode): load sessions on migration complete
* feat(kilo-vscode): ignore environment details user part when migrating sessions
* feat(cli): create project details following current backend logic
* feat(kilo-vscode): cleaner messages without environment_details parts
* feat(kilo-vscode): refactoring
* feat(kilo-vscode): refactoring and organization
* feat(kilo-vscode): method renaming to improve clarity
* feat(kilo-vscode): added session migration test strategy
* test(kilo-vscode): implement legacy migration unit tests for all migration modules
Replace todo stubs with full test implementations across messages,
parts, tools, reasoning, parser, migrate, project, and session test
files. Add new project.test.ts and session.test.ts covering project
construction, session metadata, and deterministic ID generation.
* style(kilo-vscode): apply prettier formatting across legacy migration source and test files
* fix(legacy-migration): filter out non-user/assistant entries before parsing messages
Exclude system and other non-conversational roles from the conversation
array prior to mapping, ensuring parentID references remain consistent
across imported messages when skipped entries exist.
* refactor(parts): skip part generation for non-conversational message roles
Entries with roles other than user or assistant should not produce
parts, keeping part output consistent with the messages that are
actually imported during legacy migration.
* fix(ids): use semantic kind strings to prevent part ID collisions for reasoning entries
Replace numeric indices with descriptive kind strings ("reasoning",
"provider-reasoning") in extra part ID generation to avoid collisions
between reasoning parts and regular content parts within the same
message.
Introduce `createExtraPartID` to distinguish reasoning-specific IDs
from standard sequential part IDs produced by `createPartID`.
* refactor(session-import): guard against empty worktree in legacy project import
Prevent silent misattribution of migrated sessions by rejecting empty
worktree paths before resolving the project directory. Without this
check, an empty string would fall back to the current process directory,
causing the imported session to be attached to the wrong project.
* perf(cli): cache MCP listTools results to avoid redundant RPCs per loop step (#7922)
MCP.tools() called client.listTools() for every connected server on
every agent loop iteration. With multiple servers this added hundreds
of milliseconds of overhead per step (measured ~300-750ms for remote
servers like Vercel).
Cache listTools results in a module-level Map and return them on
subsequent calls. The cache is invalidated on:
- ToolListChangedNotification from the MCP server
- Server add/connect/disconnect
- Instance dispose (config change, project switch, shutdown)
* fix(agent-manager): add retries to worktree cleanup for macOS .app bundles (#7921)
* chore: hide non-English localization files in GitHub PR diffs
Mark all non-English i18n locale files as linguist-generated=true so
they are collapsed by default in GitHub PR diffs, reducing noise when
reviewing changes that touch translation files.
* feat(vscode): add Browse Marketplace button to MCP settings
Add a "Browse Marketplace" button to the MCP Servers subtab in Agent
Behaviour settings, guiding users to discover and install MCPs from the
marketplace rather than only configuring them manually.
- Add button above the MCP server list that opens the Marketplace panel
- Add openMarketplacePanel webview message type and handler
- Add translations for all 18 supported languages
* fix(vscode): preserve project directory when opening marketplace from settings
Pass the current panel's project directory through to the marketplace
panel so project-scoped MCP installs work correctly in multi-root
workspaces. Without this, the marketplace panel would fall back to
null project directory since there is no active text editor when
clicking from a webview.
* feat(vscode): add Browse Marketplace button to agents and skills tabs
Add the Browse Marketplace button to the Agents and Skills subtabs in
Agent Behaviour settings, matching the existing button in the MCP
Servers subtab. This lets users discover marketplace items from any
of the three configuration tabs.
- Agents tab: added as a ghost button in the header row alongside
Import and Create Mode buttons
- Skills tab: added as a secondary button at the top-right, matching
the MCP tab layout
- Extracted the browse handler to the component scope for reuse
* chore: update kilo-vscode visual regression baselines
* perf(agent-manager): extend slimPart to strip heavy tool metadata for faster session switching (#7927)
Extend slim-metadata.ts to handle apply_patch, multiedit, write, and bash
tools in addition to edit. Previously only edit tool parts had their heavy
metadata stripped before IPC serialization. apply_patch parts carried full
file before/after contents per file, write parts carried the entire written
file, and bash parts carried up to 30KB of output — all serialized on every
session switch.
This reduces session-switch time by ~2x for sessions with many tool calls.
* style(legacy-migration): reformat long reasoning push call for readability
* refactor(legacy-migration): filter non-user/assistant entries before indexing
Move the role filter from inside `parseParts` to `parsePartsFromConversation`
so that message indices are derived only from the filtered list. This ensures
part message IDs stay aligned with the imported messages and skipped entries
(e.g. system role) do not cause index gaps in the generated IDs.
* fix(legacy-migration): add `_migrated_` infix to generated IDs
Prefix all migrated entity IDs (sessions, messages, parts) with a
`_migrated_` segment so they are clearly distinguishable from IDs
produced by the regular runtime. Updates tests to assert the new
prefix format.
* refactor(legacy-migration): consolidate conversation parsing into `parseSession`
Remove the intermediate `createMessages` and `createParts` async
wrappers that each independently read and parsed the conversation
file. The file is now fetched and parsed once in `parseSession`,
with the resulting conversation passed directly to
`parseMessagesFromConversation` and `parsePartsFromConversation`,
eliminating redundant I/O.
* fix(parts-builder): remove `toTextWithinMessage` and use `toText` directly
`toTextWithinMessage` was a duplicate of `toText` with identical
behavior. Replace all call sites with `toText` and delete the
redundant function.
* fix(parts-builder): skip provider-specific reasoning when explicit reasoning entry exists
Guard `isProviderSpecificReasoning` with `!isReasoning` to prevent
duplicate reasoning parts when an entry carries both a `type:
"reasoning"` field and a provider-specific reasoning field (e.g.
`reasoning_content`). The explicit reasoning entry takes precedence.
* test(legacy-migration): add coverage for edge cases in parts and tools parsing
Add unit tests for two previously uncovered scenarios:
- Task block extraction: verify only the content inside `<task>` tags
is kept when text exists outside the legacy task wrapper
- Empty tool result: verify the tool name is used as fallback output
when `tool_result` carries no readable text content
Also replace the dynamic `FileType.Directory` workaround with the
direct `vscode.FileType.Directory` enum reference, and harden
`parseFile` to throw on non-array JSON instead of silently returning
an empty array.
* fix(migration): require all items to succeed before marking migration complete
Tighten the completion condition so migration is only marked "completed"
when every item succeeds, not just when at least one does. Introduce a
dedicated "error" phase in the wizard UI so partial failures are surfaced
clearly rather than silently collapsed into the "done" state.
- `handleStartLegacyMigration`: gate `setMigrationStatus("completed")` on
`!failed && success` instead of `success` alone
- `migrate`: propagate the actual `Error.message` from a failed session
migration instead of a generic fallback string
- `MigrationWizard`: add `"error"` phase, show per-group error messages
inline, add a "Continue" button to proceed past the error screen, and
trigger `loadSessions` on completion
- `migration.css`: add `.migration-wizard__error-text` style using
`--vscode-errorForeground`
* fix(migration): defer disposeGlobal and loadSessions until after error check
Move `disposeGlobal` call inside the success-only branch so the
extension is not torn down when migration partially fails. Similarly,
defer the `loadSessions` webview message to the "Continue" button on
the error screen so sessions are only reloaded after the user
acknowledges the failure.
- `handleStartLegacyMigration`: move `disposeGlobal` inside the
`!failed && success` guard to avoid disposing on partial failure
- `MigrationWizard`: skip `loadSessions` post-message when errors are
present; send it from the "Continue" button instead
* refactor(migration): extract MigrationError component and error selector utilities
Replace inline error text rendering in MigrationWizard with a dedicated
MigrationError component that displays a structured error box with a
copy-to-clipboard action. Extract error parsing logic into reusable
selector functions in a separate module.
- Add `MigrationError` component with header, detail text, and
preformatted error code block; copy button writes error to clipboard
and shows a success toast
- Add `error-selectors.ts` with `getMigrationError` (strips error type
prefix), `getSessionMigrationError`, and `getSessionMigrationErrorDetail`
- Replace `<Show>` + raw `error-text` div in MigrationWizard sessions
row with `<MigrationError>` using the new selector-derived signals
* feat(migration): introduce view-model module for migration state derivations
Extract `ProgressEntry` type and inline derived-state functions from
`MigrationWizard` into a dedicated `migration-view-model.ts` module,
and update call sites to use the new public API.
- Add `migration-view-model.ts` exporting `ProgressEntry`, `getGroupStatus`,
`getGroupMessage`, `getSuccessCount`, `getTotalCount`,
`getCurrentSessionError`, and `getCurrentSessionErrorDetail`
- Remove duplicated logic from `MigrationWizard` and delegate to the
new module, keeping the component focused on rendering concerns
- Consolidate the `getGroupMessage` resolver used by error selectors so
it is derived from a single source of truth
* refactor(legacy-migration): split shared types into dedicated modules
Move `MigrationResultItem` out of `legacy-types.ts` into a new
`migration-types.ts` module, and introduce a `MigrationFailure`
typed error structure in `errors/migration-failure.ts`.
- Add `migration-types.ts` exporting `MigrationResultItem` to separate
runtime migration result types from legacy configuration types
- Add `errors/migration-failure.ts` defining `MigrationFailure` and
`MigrationFailureKind` for structured error classification
- Remove `MigrationResultItem` from `legacy-types.ts` to eliminate
the coupling between legacy config shapes and migration output types
- Update `migration-service.ts` import to reference the new module
* feat(errors): add migration error normalization utilities
Introduce `migration-error.ts` with `normalizeMigrationError` and
`getMigrationErrorMessage` to convert arbitrary thrown values into
typed `MigrationFailure` objects.
- Classify errors by kind (`http`, `sdk`, `generic`, `unknown`) based
on the presence of `status`, `data`, `body`, or `Error` instance
- Extract human-readable message from `message`, `body`, `data`, or
raw string fallback, defaulting to "Unknown migration error"
- Compose a `detail` string from status code, body, and data fields
when they differ from the primary message
- Propagate nested `cause` text for additional diagnostic context
* refactor(migrate): wrap migrate return in typed Result and normalize errors
* chore(migration): inline view-model helpers and remove extracted modules
Remove the separately extracted `migration-view-model.ts`, `error-selectors.ts`,
`MigrationError.tsx`, and `MigrationFailure` type in favour of inlined logic
directly in `MigrationWizard` and a simplified `getMigrationErrorMessage`
utility.
- Delete `MigrationFailure` type and `normalizeMigrationError`; replace with
a flat `getMigrationErrorMessage` function
- Inline `ProgressEntry` interface, group-status logic, and count helpers
into `MigrationWizard`
- Replace `MigrationError` component with inline clipboard copy handler
using `showToast`
- Simplify session migrate result shape: drop structured `error` field,
surface plain `message` string instead
* feat(i18n): add migration error copy-to-clipboard translations across all locales
* style(migration): add error box and copy button styles for migration wizard
Add CSS for the error display box and copy-to-clipboard button in the
migration wizard, including layout, typography, theming via VSCode CSS
variables, hover/active/focus-visible states, and high-contrast border
overrides. Also align wizard items to flex-start for multi-line content.
* fix: merge conflict issues resolved
* fix: formatting
* i18n(migration): add localized strings for session migration progress and errors
* fix:formatting
* feat(session-import): remove temporary test error from project import
* fix(vscode): remove unused migration progress translations
* fix: formatting
* fix(migration): track individual session progress instead of grouping all sessions
Replace the single "Chat sessions" progress item with per-session
tracking, passing the session ID directly to onProgress calls and
mapping each session to its own pending item in the wizard's progress
list.
* fix: formatting
---------
Co-authored-by: Marius <marius@kilocode.ai>
Co-authored-by: kiloconnect[bot] <240665456+kiloconnect[bot]@users.noreply.github.com>
Co-authored-by: Mark IJbema <mark@kilocode.ai>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
* feat(vscode): detect sessions in old extension
* feat(vscode): added session types
* feat(cli): add endpoints for inserting a session
* feat(vscode): use session types from sdk for parsing
* feat(kilo-vscode): parsing session project
* feat(vscode): parse legacy session details
* feat(kilo-vscode): refactoring
* feat(vscode): parse legacy session messages
* feat(vscode): parse legacy session parts
* feat(vscode): added testing files and id creation
* feat(vscode): merge tool parts on imported session
* feat(vscode): session message part refactoring
* feat(vscode): cover provider-specific reasoning parts on session migration
* feat(kilo-vscode): handle completion result parts
* feat: type hardening on session migration
* feat(kilo-vscode): create tool part correctly
* feat(kilo-vscode): adjust session part processing
* feat(vscode): persist migrated session in kilo.db
* feat(vscode): add error handling to session migration
* feat(vscode): add directory field to session migration
* feat(kilo-vscode): display completion check on session migration
* feat(vscode): prevent migrating sessions twice
* feat(cli): use correct project id when inserting migrated session
* feat(kilo-vscode): Use correct project id for session migration
* feat(kilo-vscode): load sessions on migration complete
* feat(kilo-vscode): ignore environment details user part when migrating sessions
* feat(cli): create project details following current backend logic
* feat(kilo-vscode): cleaner messages without environment_details parts
* feat(kilo-vscode): refactoring
* feat(kilo-vscode): refactoring and organization
* feat(kilo-vscode): method renaming to improve clarity
* feat(kilo-vscode): added session migration test strategy
* test(kilo-vscode): implement legacy migration unit tests for all migration modules
Replace todo stubs with full test implementations across messages,
parts, tools, reasoning, parser, migrate, project, and session test
files. Add new project.test.ts and session.test.ts covering project
construction, session metadata, and deterministic ID generation.
* style(kilo-vscode): apply prettier formatting across legacy migration source and test files
* fix(legacy-migration): filter out non-user/assistant entries before parsing messages
Exclude system and other non-conversational roles from the conversation
array prior to mapping, ensuring parentID references remain consistent
across imported messages when skipped entries exist.
* refactor(parts): skip part generation for non-conversational message roles
Entries with roles other than user or assistant should not produce
parts, keeping part output consistent with the messages that are
actually imported during legacy migration.
* fix(ids): use semantic kind strings to prevent part ID collisions for reasoning entries
Replace numeric indices with descriptive kind strings ("reasoning",
"provider-reasoning") in extra part ID generation to avoid collisions
between reasoning parts and regular content parts within the same
message.
Introduce `createExtraPartID` to distinguish reasoning-specific IDs
from standard sequential part IDs produced by `createPartID`.
* refactor(session-import): guard against empty worktree in legacy project import
Prevent silent misattribution of migrated sessions by rejecting empty
worktree paths before resolving the project directory. Without this
check, an empty string would fall back to the current process directory,
causing the imported session to be attached to the wrong project.
* perf(cli): cache MCP listTools results to avoid redundant RPCs per loop step (#7922)
MCP.tools() called client.listTools() for every connected server on
every agent loop iteration. With multiple servers this added hundreds
of milliseconds of overhead per step (measured ~300-750ms for remote
servers like Vercel).
Cache listTools results in a module-level Map and return them on
subsequent calls. The cache is invalidated on:
- ToolListChangedNotification from the MCP server
- Server add/connect/disconnect
- Instance dispose (config change, project switch, shutdown)
* fix(agent-manager): add retries to worktree cleanup for macOS .app bundles (#7921)
* chore: hide non-English localization files in GitHub PR diffs
Mark all non-English i18n locale files as linguist-generated=true so
they are collapsed by default in GitHub PR diffs, reducing noise when
reviewing changes that touch translation files.
* feat(vscode): add Browse Marketplace button to MCP settings
Add a "Browse Marketplace" button to the MCP Servers subtab in Agent
Behaviour settings, guiding users to discover and install MCPs from the
marketplace rather than only configuring them manually.
- Add button above the MCP server list that opens the Marketplace panel
- Add openMarketplacePanel webview message type and handler
- Add translations for all 18 supported languages
* fix(vscode): preserve project directory when opening marketplace from settings
Pass the current panel's project directory through to the marketplace
panel so project-scoped MCP installs work correctly in multi-root
workspaces. Without this, the marketplace panel would fall back to
null project directory since there is no active text editor when
clicking from a webview.
* feat(vscode): add Browse Marketplace button to agents and skills tabs
Add the Browse Marketplace button to the Agents and Skills subtabs in
Agent Behaviour settings, matching the existing button in the MCP
Servers subtab. This lets users discover marketplace items from any
of the three configuration tabs.
- Agents tab: added as a ghost button in the header row alongside
Import and Create Mode buttons
- Skills tab: added as a secondary button at the top-right, matching
the MCP tab layout
- Extracted the browse handler to the component scope for reuse
* chore: update kilo-vscode visual regression baselines
* perf(agent-manager): extend slimPart to strip heavy tool metadata for faster session switching (#7927)
Extend slim-metadata.ts to handle apply_patch, multiedit, write, and bash
tools in addition to edit. Previously only edit tool parts had their heavy
metadata stripped before IPC serialization. apply_patch parts carried full
file before/after contents per file, write parts carried the entire written
file, and bash parts carried up to 30KB of output — all serialized on every
session switch.
This reduces session-switch time by ~2x for sessions with many tool calls.
* style(legacy-migration): reformat long reasoning push call for readability
* refactor(legacy-migration): filter non-user/assistant entries before indexing
Move the role filter from inside `parseParts` to `parsePartsFromConversation`
so that message indices are derived only from the filtered list. This ensures
part message IDs stay aligned with the imported messages and skipped entries
(e.g. system role) do not cause index gaps in the generated IDs.
* fix(legacy-migration): add `_migrated_` infix to generated IDs
Prefix all migrated entity IDs (sessions, messages, parts) with a
`_migrated_` segment so they are clearly distinguishable from IDs
produced by the regular runtime. Updates tests to assert the new
prefix format.
* refactor(legacy-migration): consolidate conversation parsing into `parseSession`
Remove the intermediate `createMessages` and `createParts` async
wrappers that each independently read and parsed the conversation
file. The file is now fetched and parsed once in `parseSession`,
with the resulting conversation passed directly to
`parseMessagesFromConversation` and `parsePartsFromConversation`,
eliminating redundant I/O.
* fix(parts-builder): remove `toTextWithinMessage` and use `toText` directly
`toTextWithinMessage` was a duplicate of `toText` with identical
behavior. Replace all call sites with `toText` and delete the
redundant function.
* fix(parts-builder): skip provider-specific reasoning when explicit reasoning entry exists
Guard `isProviderSpecificReasoning` with `!isReasoning` to prevent
duplicate reasoning parts when an entry carries both a `type:
"reasoning"` field and a provider-specific reasoning field (e.g.
`reasoning_content`). The explicit reasoning entry takes precedence.
* test(legacy-migration): add coverage for edge cases in parts and tools parsing
Add unit tests for two previously uncovered scenarios:
- Task block extraction: verify only the content inside `<task>` tags
is kept when text exists outside the legacy task wrapper
- Empty tool result: verify the tool name is used as fallback output
when `tool_result` carries no readable text content
Also replace the dynamic `FileType.Directory` workaround with the
direct `vscode.FileType.Directory` enum reference, and harden
`parseFile` to throw on non-array JSON instead of silently returning
an empty array.
* feat(legacy-migration): remove session migration tracking from global state
Drop the `readSessionsToMigrate` filter and the post-migration
`globalState` flag that tracked which sessions had already been
migrated. Detection now returns all sessions found in global storage
without filtering by prior migration status.
* fix(session-import): skip duplicate session imports and child data
Check for an existing session row before inserting and return a
`skipped: true` flag when the session already exists. The migration
client honours this flag to bypass message and part imports, avoiding
redundant work when migration is re-run.
Replace `onConflictDoUpdate` with `onConflictDoNothing` on the session
insert path to align with the new pre-check behaviour. Propagate the
optional `skipped` field through the result types and generated SDK
types for project, session, message, and part responses.
* refactor(parts-builder): mark legacy system error text parts as ignored
Tag text parts whose content begins with `[ERROR]` with `ignored: true`
and a `legacy-system-error` metadata source. Introduces a new
`isLegacySystemErrorText` utility in `parts-util` to encapsulate the
detection logic.
* feat(parts-util): extract feedback text from tool_result content into visible parts
Parse `<feedback>` tags embedded in `tool_result` content and emit a
dedicated text part so user feedback is surfaced as a visible message
rather than buried inside tool result data.
Introduces `getFeedbackText` utility to extract and trim the inner
content of `<feedback>` blocks from arbitrary input.
* fix: formatting
---------
Co-authored-by: Marius <marius@kilocode.ai>
Co-authored-by: kiloconnect[bot] <240665456+kiloconnect[bot]@users.noreply.github.com>
Co-authored-by: Mark IJbema <mark@kilocode.ai>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Swap order in KilocodePaths.skillDirectories() so global dirs (~/.kilo,
~/.kilocode, VSCode storage) are pushed first, followed by project-level
dirs. Since addSkill uses last-write-wins, project-level skills now
correctly take precedence over global skills.
Fixes#7886
When a provider emits tool-call without a preceding tool-input-start
event, the tool call was silently dropped causing conversation history
corruption and 400 errors on subsequent API requests.
This change creates a fallback tool part in the tool-call handler when
tool-input-start was never emitted, ensuring the tool call is persisted
and included in conversation history.
On Windows, VS Code's uri.fsPath returns lowercase drive letters (e.g. c:\...)
while Filesystem.resolve() canonicalises them to uppercase (C:\...) via
realpathSync.native. Sessions are stored with the resolved path but queried
with the raw fsPath, so SQLite's binary equality check never matches and no
sessions are shown after restart.
Apply Filesystem.resolve() to the directory filter in Session.list and
Session.listGlobal so both sides use the same canonical form.
Kirill Kalishev