root/fusion
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity
11,279 Commits 1 Branch 0 Tags
9a43aa1d240d850db65448097201d883b5bd5595
Commit Graph

11279 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
gsxdsm
9a43aa1d24 FN-7951: harden runGenerationWithTimeout abort cancellation across planning surfaces 
Ensures aborted AI generation (timeout, user-stop, displacement, retries) actually tears down the in-flight agent session instead of only rejecting the Promise.race waiter, since provider SDKs may ignore AbortSignal.

- Add a once-only onAbort teardown hook to GenerationGuard, invoked for timeout, user-stop, and displaced abort causes so consumers can dispose their in-flight session exactly once.
- Give planning's local generation runner (runGenerationWithTimeout) the same guaranteed once-only abortTeardown for timeout, user-stop, displacement, stuck, and loop aborts, replacing the ad hoc dispose-on-timeout-only logic.
- Forward the AbortSignal into planning's history-replay prompt, turn prompts, and JSON-parse-retry prompts, and short-circuit with createAbortError() when the signal is already aborted before/after each prompt call.
- Wire subtask-breakdown's onTimeout/onUserStop handlers to the new onAbort hook instead of disposing the agent directly, keeping teardown centralized in the guard.
- Add GenerationInProgressError / TargetGenerationInProgressError handling in mission-routes to return 409 Conflict instead of a generic 500 when a generation is already running.
- Extend mission-interview and milestone-slice-interview generation paths with matching abort-forwarding and teardown behavior, plus new/expanded tests covering cancellation across timeout, user-stop, displacement, and retry paths.
- Add a patch changeset documenting the fix for @runfusion/fusion.

Files changed:
 .changeset/harden-generation-abort.md              |   7 ++
 .../src/__tests__/ai-session-timeout.test.ts       |  41 +++++--
 .../__tests__/milestone-slice-interview.test.ts    |  72 ++++++++++++-
 .../src/__tests__/mission-interview.test.ts        |  64 ++++++++++-
 .../planning-generation-cancellation.test.ts       |  82 ++++++++++++++
 .../src/__tests__/subtask-breakdown.test.ts        |  21 +++-
 packages/dashboard/src/ai-session-timeout.ts       |  33 +++++-
 .../dashboard/src/milestone-slice-interview.ts     | 120 +++++++++++++++++++--
 packages/dashboard/src/mission-interview.ts        | 119 ++++++++++++++++++--
 packages/dashboard/src/mission-routes.ts           |  12 +++
 packages/dashboard/src/planning.ts                 |  70 +++++++++---
 packages/dashboard/src/subtask-breakdown.ts        |  10 +-
 12 files changed, 589 insertions(+), 62 deletions(-)

Fusion-Task-Id: FN-7951
Fusion-Task-Lineage: debcd6a9-f54e-4ef3-87e1-4f06be0b5f64
Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 13:58:35 -07:00
gsxdsm
6e0fde860c FN-7949: fix deleted planning-mode session resurrecting after in-flight generation completes 
Adds a bounded-TTL delete tombstone to AiSessionStore so a straggling post-delete generation write can never resurrect a session the user explicitly deleted.

- AiSessionStore now records a 10-minute delete tombstone (id -> deletion timestamp) in delete(), deleteByIdAndType(), and bulk cleanup paths (cleanupOld/cleanupStaleSessions/emitDeletedSessions).
- upsert() checks the tombstone first and drops (no-ops) any write for a tombstoned id without touching SQLite or emitting ai_session:updated, fixing the root cause once in the shared store rather than per-producer (planning.ts, subtask-breakdown.ts, mission-interview.ts, milestone-slice-interview.ts).
- Tombstone entries are pruned lazily on check and piggyback on the existing cleanupStaleSessions() cadence so the in-memory map cannot grow unbounded.
- Adds a changeset (patch) documenting the user-facing fix.
- Updates docs/architecture.md and docs/storage.md with the new "AI session delete tombstones" behavior.
- Adds regression tests covering the tombstone guard in ai-session-store.test.ts and routes-planning.test.ts.

Files changed:
 .changeset/fn-7949-ai-session-delete-tombstone.md  |   7 +
 docs/architecture.md                               |   2 +-
 docs/storage.md                                    |  12 +-
 packages/dashboard/src/__tests__/ai-session-store.test.ts | 145 +++++++++++++++
 packages/dashboard/src/__tests__/routes-planning.test.ts  | 200 ++++++++++++++++++++-
 packages/dashboard/src/ai-session-store.ts         |  83 +++++++++
 6 files changed, 446 insertions(+), 3 deletions(-)

Fusion-Task-Id: FN-7949

Fusion-Task-Lineage: 8e509dae-0cc5-46cd-9c4b-9048cfda56d3

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 13:10:53 -07:00
gsxdsm
1ff83a2735 chore(release): v0.60.0 
Version bump via changesets.
 2026-07-13 10:32:12 -07:00
gsxdsm
4e7e013d6f FN-7947: add Plan action to context menu for pre-execution task cards 
Adds a Plan action to Board/List task context menus so triage/hold/intake cards can jump straight into Planning Mode without duplicating a task.

- Add `onPlan` handler and `isPreExecutionHoldColumn` gate to `TaskContextMenu` so Plan only appears for pre-execution (triage/intake/hold) columns, and only when a host wires the handler
- Wire the Plan action through `Board.tsx`, `Column.tsx`, `ListView.tsx`, and `WorktreeGroup.tsx` so both board and list views expose the new menu item
- Surface the Plan entry point on `TaskCard.tsx`
- Add test coverage in `TaskContextMenu.test.tsx`, `TaskCard.test.tsx`, and `ListView.test.tsx` for the new gating/wiring behavior
- Document the new action in `docs/dashboard-guide.md`
- Add a minor changeset for `@runfusion/fusion`

Files changed:
 .changeset/fn-7947-plan-context-menu-action.md     |  7 ++
 docs/dashboard-guide.md                            | 10 ++-
 packages/dashboard/app/components/Board.tsx        | 10 ++-
 packages/dashboard/app/components/Column.tsx       |  4 +
 packages/dashboard/app/components/ListView.tsx     | 15 +++-
 packages/dashboard/app/components/TaskCard.tsx     | 24 +++++-
 packages/dashboard/app/components/TaskContextMenu.tsx   | 18 ++++
 packages/dashboard/app/components/WorktreeGroup.tsx     |  9 ++
 packages/dashboard/app/components/__tests__/ListView.test.tsx     | 21 +++++
 packages/dashboard/app/components/__tests__/TaskCard.test.tsx     | 96 ++++++++++++++++++++++
 packages/dashboard/app/components/__tests__/TaskContextMenu.test.tsx  | 32 ++++++++
 11 files changed, 236 insertions(+), 10 deletions(-)

Fusion-Task-Id: FN-7947

Fusion-Task-Lineage: 41c759a2-e76b-4771-9421-c9805c4596e5

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 10:14:22 -07:00
gsxdsm
7cc622bed2 FN-7946: auto-retry stuck Planning Mode AI generation up to 3 times 
Planning Mode now automatically retries a stuck or terminally-errored AI
generation session up to three times before falling back to the permanent
Retry/Dismiss error panel, reducing manual retries for transient failures.

- Add a bounded (MAX_PLANNING_AUTO_RETRIES = 3) client-side auto-retry that
  reuses the existing /planning/:id/retry endpoint whenever the SSE stream's
  onError, a session reload, or the stuck-session poll observes a terminal
  "error" status.
- Track the retry budget in refs (planningAutoRetryAttemptRef,
  planningAutoRetryInFlightRef) so async SSE/poll/loadSession handlers share
  a single in-flight guard, with the current attempt mirrored into state
  (isAutoRetrying/autoRetryAttempt) for the UI.
- Reset the retry budget whenever the session makes real progress (reaches
  a new question or a completed summary), and surface the permanent
  Retry/Dismiss error view once the budget is exhausted.
- Show a "Retrying... (attempt N of 3)" loading message while an automatic
  retry is in flight, distinct from the manual Retry button state.
- Fix a stuck-poll edge case where a terminal error discovered only by the
  poll (missed SSE event) after the auto-retry budget was exhausted left
  the modal spinning on "Generating next question..." forever instead of
  showing the error view.
- Document the new auto-retry behavior in docs/dashboard-guide.md and add a
  minor changeset for @runfusion/fusion.
- Extend PlanningModeModal.planning-flow.test.tsx with coverage for the
  auto-retry budget, single-flight behavior, and the poll-discovered
  terminal-error fallback.

Files changed:
 .changeset/fn-7946-planning-auto-retry.md          |   7 +
 docs/dashboard-guide.md                            |   3 +
 .../dashboard/app/components/PlanningModeModal.tsx | 339 ++++++++++++++------
 .../PlanningModeModal.planning-flow.test.tsx       | 353 ++++++++++++++++++---
 4 files changed, 567 insertions(+), 135 deletions(-)

Fusion-Task-Id: FN-7946
Fusion-Task-Lineage: 42e911dc-9639-46ab-bb4f-bc9060413140
Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 09:50:48 -07:00
gsxdsm
f0888d43c3 FN-7945: route List-view task opens through the movable popup when Open tasks as popups is on 
Extends the existing board/right-dock "Open tasks as popups" routing so ordinary List row/card and keyboard opens use the same shared movable/resizable FloatingWindow instead of the docked split-pane/mobile detail.

- Add openMobileTasksInPopup prop to ListView, threaded through App -> MainContent -> ListView (dashboard/types.ts)
- handleRowClick routes to onPopOut (popOutTaskDetail) when the setting is on, on both desktop split-pane and mobile/tablet single-pane; docked behavior is preserved when the setting is off
- Restore Enter/Space keyboard activation on list rows to invoke the same handleRowClick path, alongside existing context-menu key handling
- Update docs/dashboard-guide.md and docs/settings-reference.md to describe List row/card opens as part of the popup routing surface, and refresh the Appearance settings help copy/FNXC comment accordingly
- Add changeset (.changeset/fn-7945-list-view-task-popup.md, minor) describing the user-facing behavior
- Extend ListView.test.tsx coverage for the new popup routing and restored keyboard activation

Files changed:
 .changeset/fn-7945-list-view-task-popup.md         |  7 ++
 docs/dashboard-guide.md                            |  4 +-
 docs/settings-reference.md                         |  2 +-
 packages/dashboard/app/App.tsx                     |  1 +
 packages/dashboard/app/components/ListView.tsx     | 48 +++++++++----
 .../app/components/__tests__/ListView.test.tsx     | 80 +++++++++++++++++++++-
 .../app/components/dashboard/MainContent.tsx       |  2 +
 .../dashboard/app/components/dashboard/types.ts    |  1 +
 .../settings/sections/AppearanceSection.tsx        |  4 +-
 9 files changed, 127 insertions(+), 22 deletions(-)

Fusion-Task-Id: FN-7945

Fusion-Task-Lineage: 784cb4ee-c493-4ace-bf8b-0e3dbaaef9a3

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 09:27:29 -07:00
gsxdsm
7246df22f6 FN-7944: add setting to keep task popups attached to their Board/List view 
Adds an opt-in project setting so open task-detail popups stay attached to the Board or List view where they were opened, instead of floating over every main-content view.

- New project setting taskPopupsBoardListOnly (default: off) in settings-schema.ts and ProjectSettings type, with default preserved via settings-defaults tests.
- usePoppedOutTasks now stores each popup's originating TaskView alongside its task snapshot (PoppedOutTaskEntry), keeping legacy tasks output for existing callers.
- App.tsx adds isTaskPopupVisibleForView() gating helper and filters popped-out entries to the current view for rendering/keyboard-close handling, while hidden popups remain mounted in hook state (not cleared) so switching back to the originating view restores them with shared persisted geometry.
- Settings -> Appearance gets a new "Keep task popups on their Board/List view" checkbox (AppearanceSection.tsx) with i18n strings and updated settings search text in SettingsModal.
- Documentation updated in docs/dashboard-guide.md and docs/settings-reference.md to describe the render-only hide/restore behavior.
- New/updated tests: App.taskPopupViewGating.test.tsx, usePoppedOutTasks.test.ts, AppearanceSection.test.tsx, settings-default-descriptions.test.tsx, settings-defaults.test.ts.

Files changed:
 docs/dashboard-guide.md                            |   5 +-
 docs/settings-reference.md                         |   1 +
 .../core/src/__tests__/settings-defaults.test.ts   |  13 +++
 packages/core/src/settings-schema.ts               |   5 +
 packages/core/src/types.ts                         |   7 ++
 packages/dashboard/app/App.tsx                     |  49 +++++++--
 .../app/__tests__/App.taskPopupViewGating.test.tsx | 113 +++++++++++++++++++++
 .../dashboard/app/components/SettingsModal.tsx     |   3 +-
 .../settings/sections/AppearanceSection.tsx        |   8 ++
 .../sections/__tests__/AppearanceSection.test.tsx  |  21 ++++
 .../settings-default-descriptions.test.tsx         |   1 +
 .../app/hooks/__tests__/usePoppedOutTasks.test.ts  |  14 +++
 packages/dashboard/app/hooks/useAppSettings.ts     |   4 +
 packages/dashboard/app/hooks/usePoppedOutTasks.ts  |  27 +++--
 packages/i18n/locales/en/app.json                  |   2 +
 15 files changed, 255 insertions(+), 18 deletions(-)

Fusion-Task-Id: FN-7944
Fusion-Task-Lineage: 4b8ced0e-1853-429f-8482-163821a35ae6
Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 08:58:51 -07:00
gsxdsm
6b78633b07 FN-7943: keep Quick Chat open when portaled model/thinking-level dropdowns are clicked 
Quick Chat's outside-pointer dismissal now recognizes body-portaled dropdown menus (model, thinking-level, agent, dependency, node, priority) as part of the panel instead of treating them as outside clicks.

- Extend FloatingWindow's outside-pointerdown safe-surface selector to include the portaled dropdown classes used by model combobox, model nested menu, dependency, node picker, agent picker, and priority picker menus
- Add regression tests covering pointerdown on each portaled dropdown surface and on a child element inside a portaled dropdown, asserting onClose is not called
- Update dashboard-guide docs to describe that these portal dropdowns are treated as part of the Quick Chat panel for outside-click purposes

Files changed:
 docs/dashboard-guide.md                            |  2 +-
 .../dashboard/app/components/FloatingWindow.tsx    | 19 +++++++-
 .../components/__tests__/FloatingWindow.test.tsx   | 50 ++++++++++++++++++++++
 3 files changed, 69 insertions(+), 2 deletions(-)

Fusion-Task-Id: FN-7943

Fusion-Task-Lineage: fa91bd43-241c-48b0-8858-16521f383784

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 08:15:25 -07:00
gsxdsm
c3c3861efa fix: stale TRANSITIVE_EXTERNALS + dashboard dist/client clean + cross-worker build lock (round 12) (#2065) 
## Summary

Fixes shard 3 failures from runs 29258546612 + 29259574946 (FN-7936
drift).

## Fixes

### `package-config.test.ts` — stale TRANSITIVE_EXTERNALS entry
FN-7936 aliased `@fusion/core` to a runtime shim in bundled plugin
outputs; it's no longer a tsup external. Removed the stale allowlist
entry.

### `bundle-output.test.ts` — stale dashboard client hash ENOENT
**Root cause:** Two test files (`bundle-output.test.ts` +
`extension-integration.test.ts`) call
`buildCliWithRealDashboardAssets()` which triggers concurrent vite/tsup
builds. Vitest runs them in parallel (`pool: "forks"`, `fileParallelism:
true`). Without coordination, two builds clean and write `dist/client`
simultaneously, causing `ENOENT` on content-hashed chunk files.

**Fix (3 parts):**
1. **`workspace-tools.ts buildDashboardClient`** — `rm dist/client`
before vite build. Prevents stale content-hash references from previous
builds.
2. **`bundle-output-helpers.ts`** — atomic `mkdirSync` file lock around
`buildCliWithRealDashboardAssets()`. Winner builds; losers poll with
`Atomics.wait`, then re-check `hasBuiltDashboardAssets()`. On timeout,
**throws** (never builds without owning the lock).
3. Lock uses `Atomics.wait(new Int32Array(new SharedArrayBuffer(4)), 0,
0, 500)` for sync sleep — no child process spawning.

## Verification
- Gate: exit 0 ✅

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Bug Fixes**
* Improved dashboard asset builds by removing stale files before
rebuilding.
* Prevented concurrent builds from producing incomplete or corrupted
dashboard assets.
* Added safeguards to detect stalled asset builds and fail with clearer
errors.

* **Tests**
* Updated package validation checks to reflect current runtime bundling
behavior.
  * Improved reliability of CLI build-related test execution.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
 2026-07-13 08:14:20 -07:00
gsxdsm
d4001ab0ee feat: make merger AI model configurable under Global and Project Models 
Add a dedicated merger model lane (project + global provider/model/thinking) so merge-agent sessions no longer share only the default model, without inheriting executor/planner/reviewer lanes.
 2026-07-13 08:10:56 -07:00
gsxdsm
e35620c9aa FN-7939: supervise heartbeat timer-audit interval and bound non-advancing zombie re-arms 
Fixes agents silently going stale for hours even though the heartbeat repair audit process was running.

- HeartbeatTriggerScheduler now runs an independent watchdog (armTimerAuditWatchdog/checkTimerAuditLiveness) that tracks the audit loop's last-run timestamp and re-arms + immediately re-runs the 60s audit interval if it goes stale beyond a bounded multiple of the cadence, so a silently dropped audit driver self-heals instead of leaving active agents unrepaired for hours.
- Tracks consecutive non-advancing zombie-timer re-arms per agent (nonAdvancingRearmState) and escalates once the count crosses a threshold, recording consecutiveNonAdvancingRearms/nonAdvancingEscalated in agent.metadata.heartbeatTimerRepair and logging reason=heartbeat-rearm-nonadvancing-escalated instead of silently churning the same zombie-timer-rearmed repair forever.
- Clears non-advancing rearm state on unregister, non-eligible agents, paused settings, and stale-run-reap skip paths so tracking never leaks stale per-agent counters.
- Watchdog and its interval handle are armed in start() and cleared in stop() alongside the existing audit interval.
- Adds a changeset (patch) describing the fix, and updates docs/agents.md and docs/architecture.md to document the FN-7939 audit watchdog and non-advancing escalation behavior.
- Adds heartbeat-scheduler.test.ts coverage for watchdog re-arm/liveness and non-advancing escalation.

Files changed:
 .changeset/fn-7939-heartbeat-audit-supervision.md  |   7 +
 docs/agents.md                                     |   8 +-
 docs/architecture.md                               |   1 +
 .../src/__tests__/heartbeat-scheduler.test.ts      | 209 +++++++++++++++++++++
 packages/engine/src/agent-heartbeat.ts             | 128 ++++++++++++-
 5 files changed, 341 insertions(+), 12 deletions(-)

Fusion-Task-Id: FN-7939
Fusion-Task-Lineage: 9fa90240-4333-4588-b595-aef3811b1524
Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 07:51:07 -07:00
gsxdsm
316d4fa034 FN-7941: anchor execute-requeue loop guard to monotonic terminal-step progress 
Hardens the FN-7863 execute-node self-requeue loop guard so residual execute_loop_stall cases (#2043/#2045/#2046/#2047) can no longer reset the loop counter forever via non-terminal signature drift.

- Change buildExecuteRequeueLoopSignature to track terminal step count (done/skipped) plus total step count instead of raw currentStep + every step status, so pending/in-progress oscillation no longer produces a "new" signature each cycle.
- Add buildExecuteRequeueLoopHighWaterSignature, which derives current terminal-step progress via the shared signature parser (parseExecuteRequeueLoopProgressSignature) and only resets the streak on monotonic forward progress, keeping a high-water mark across cycles so decreases/oscillation below the high-water still count toward exhaustion.
- Update executor.ts's execute self-requeue dispatch path to use the new high-water helper when deciding whether to reset (1) or increment executeRequeueLoopCount, replacing the previous raw signature-equality check.
- Extend execute-requeue-loop-guard.test.ts with regression coverage: a drifting-signature case that oscillates step order/status with no terminal progress (still terminalizes at MAX_EXECUTE_REQUEUE_LOOP_CYCLES), a done/in-progress oscillation case bounded after the high-water stops increasing, and an updated "real progress never terminalizes" case driven by genuine monotonic done-step advancement.
- Update docs/architecture.md's FN-7863/FN-7926 self-healing notes to describe the new terminal-step high-water signature and cross-reference FN-7941.

Files changed:
 docs/architecture.md                               |  4 +-
 .../execute-requeue-loop-guard.test.ts             | 83 +++++++++++++++++++++-
 packages/engine/src/executor.ts                    | 54 ++++++++++++--
 3 files changed, 130 insertions(+), 11 deletions(-)

Fusion-Task-Id: FN-7941

Fusion-Task-Lineage: cbf1e536-d29b-40da-bdd8-8c34d8d6b1ca

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 07:46:54 -07:00
gsxdsm
cf1b33bf5e FN-7942: index Project Models chat-default settings in Settings search 
Extends the shared Settings search index so queries like "chat" surface the Project Models section's Direct-chat default fields, which previously had no matching searchable terms.

- Add chat-related searchableText keywords (chat, new chat, chat default, chat model, chat agent, etc.) to the project-models SETTINGS_SECTIONS entry
- Add searchableKeys covering the i18n chat-default labels (chatHeading, chatDescription, chatNewSessionMode*, chatDefaultKind/Model/Agent) so translated labels are also indexed
- Export SettingsSection type plus normalizeSettingsSearchText/sectionMatchesSettingsSearch/filterSettingsSectionsForSearch/SETTINGS_SECTIONS from SettingsModal.tsx for direct unit testing
- Add SettingsModal.search.test.ts covering chat-query matches and a negative case (unrelated Remote Access term does not match)
- Add changeset fn-7942-settings-search-chat.md (patch, fix category)

Files changed:
 .changeset/fn-7942-settings-search-chat.md         |  7 +++
 .../dashboard/app/components/SettingsModal.tsx     | 50 +++++++++++++++++++---
 .../__tests__/SettingsModal.search.test.ts         | 41 ++++++++++++++++++
 3 files changed, 92 insertions(+), 6 deletions(-)

Fusion-Task-Id: FN-7942

Fusion-Task-Lineage: fa044908-7ada-4334-a789-6d2bf08afdb8

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 07:44:12 -07:00
gsxdsm
281bb05161 FN-7936: alias @fusion/core to a runtime shim in bundled plugin outputs 
Fix bundled example plugins (dependency-graph, grok-runtime, roadmap, acp-runtime, compound-engineering) crashing on enable with "Cannot find package '@fusion/core'" by aliasing the private import to a self-contained runtime shim during CLI bundling.

- packages/cli/tsup.config.ts: drop @fusion/core from bundlePluginEntry's external list and alias it to the existing pluginSdkCoreRuntimeShim so bundled.js no longer references the private workspace package at runtime
- packages/cli/src/__tests__/bundle-output.test.ts: add a regression test asserting every staged bundled plugin's bundled.js contains no bare @fusion/core import/reference
- docs/PLUGIN_AUTHORING.md: document that bundled.js outputs must be self-contained and must not leak private @fusion/* workspace imports
- .changeset/fn-7936-bundled-plugin-fusion-core-external.md: add a patch changeset for @runfusion/fusion describing the fix

Files changed:
 .changeset/fn-7936-bundled-plugin-fusion-core-external.md |  7 +++++
 docs/PLUGIN_AUTHORING.md                                  |  3 +++
 packages/cli/src/__tests__/bundle-output.test.ts          | 30 ++++++++++++++++++++++
 packages/cli/tsup.config.ts                               |  9 +++++--
 4 files changed, 47 insertions(+), 2 deletions(-)

Fusion-Task-Id: FN-7936

Fusion-Task-Lineage: a8a391b2-9441-4a7c-92bc-f1675e1a8a0d

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 07:32:50 -07:00
gsxdsm
a81486559e fix: dashboard lucide/CSS/SettingsModal/mcp-helper fixes + expand mock-completeness guard to dashboard (round 11) (#2049) 
## Summary

Fixes all remaining full-suite failures from run 29231108919
(FN-7923–7931 drift) + expands the structural mock-completeness guard to
cover dashboard tests.

## Fixes

### Dashboard (shard 4) — 4 files
- **`TaskCard.cli-states.test.tsx`** — Proxy lucide-react mock needed
`has`/`getOwnPropertyDescriptor` traps; TaskCard now imports
`priorityIndicator.tsx` which reads `ArrowDown` at module-init. Vitest
validates ESM named exports via `in`/descriptor, not `get`. Also added
`useToast` mock.
- **`SettingsModalNodeRouting.test.tsx`** — Pass
`initialSection="node-routing"` (it's in
`ADVANCED_SETTINGS_SECTION_IDS`, nav hidden by default).
- **`styles-css-rgba-tokenization.test.ts`** — Removed stale
`.settings-sidebar` color-mix expectation (FN-7825 made it
structural-only).
- **`mcp-helper-forwarding.test.ts`** — Added
`resolvePlanningThinkingLevel` to `@fusion/engine` mock (insight
extraction calls it before MCP forwarding).

### Structural guard expansion
- **`.tsx` blind spot fixed** — `collectTs` and test file filter now
include `.tsx` files
- **Shorthand property extraction** — key extractor now matches both
`key: value` and `key,` (shorthand)
- **Convention mapping** — `.test.tsx → .tsx` source resolution added
- **Dashboard test coverage** — `@fusion/engine` barrel check now scans
`packages/dashboard/src/__tests__/`
- **7 latent mock gaps completed** — `pr-conflict-resolver`,
`project-pause-resume-routes`, `routes-approval-sandbox-provisioning`,
`routes-approval`, `routes-worktrunk`, `session-reconnect`,
`setup-routes`

### Engine (shards 1+2) — zero real failures
All 3 failing files are local-only (`@agentclientprotocol/sdk` + pi-ai
staleness). CI resolves them from lockfile.

## Verification
- Gate (with expanded guard): exit 0 ✅
- Dashboard (5 non-local files): 36/36 passed ✅
- 3 files skipped locally (`@agentclientprotocol/sdk`) — CI will verify

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Tests**
* Improved test reliability by completing module mocks across dashboard
and routing scenarios.
* Updated settings and task card test coverage to reflect current UI
behavior.
* Enhanced mock validation to cover additional test files, TypeScript
React files, and shorthand exports.
* Prevented failures related to missing providers, engine helpers, and
planning configuration.


<!-- end of auto-generated comment: release notes by coderabbit.ai -->
 2026-07-13 07:10:14 -07:00
gsxdsm
502c4c132f chore(release): v0.59.0 
Version bump via changesets.
 2026-07-13 01:23:43 -07:00
gsxdsm
edfe57c8e0 feat: AI Highlights and engagement X draft on release 
Claude-author Highlights plus an engagement-oriented tweet (≤280) during changelog distillation, with scheme-free links, Fusion version openers that drop .0 patch, and a soft deterministic fallback when Claude is offline.
 2026-07-13 01:21:33 -07:00
gsxdsm
b487f4542d fix: raise TaskCard Actions control onto header centerline 
Nudge the text+chevron Actions/Send-back chip up by a quarter space-xs so
it no longer sits optically below the ⋯ menu and size badge on the locked
chip-height row.
 2026-07-13 01:06:01 -07:00
gsxdsm
0b60fe9548 fix: align TaskCard right header cluster with task-id baseline 
Lock .card-id and .card-header-actions to the chip height so the mobile
28px ⋯ touch target cannot stretch the row and sink Actions/size below
FN-####. Cancel residual menu layout growth and extend badge-wrap
regression coverage for the shared locked-row contract.
 2026-07-13 01:03:06 -07:00
gsxdsm
cf58d7234d FN-7933: align mobile task-card header controls to one centerline 
Follow-up to FN-7928: fix vertical alignment of the Send-back/Actions trigger, menu button, and size badge in the mobile task-card header actions.

- Set line-height: 1 on .card-menu-btn, .card-size-badge, and .card-send-back-btn so their text/icon baselines match instead of drifting from default line-height.
- Add align-items: center to .card-header-actions at the mobile breakpoint so Send back, menu, and size badge share one optical vertical centerline.
- Add padding-block to .card-size-badge to keep the badge's rendered height consistent with the neighboring controls after the line-height fix.
- Add regression tests (TaskCard.badge-wrap.test.tsx) asserting the mobile header-actions rule set (min-height, align-items, gap) and per-control line-height/padding declarations, plus an awaiting-user-input coverage case exercising the send-back/menu/size-badge centerline together.

Files changed:
 packages/dashboard/app/components/TaskCard.css     | 16 ++++
 .../__tests__/TaskCard.badge-wrap.test.tsx         | 92 ++++++++++++++++++++++
 2 files changed, 108 insertions(+)

Fusion-Task-Id: FN-7933
Fusion-Task-Lineage: e3d52b36-2446-4cdf-8faf-5c94b0f52786
Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 00:53:48 -07:00
gsxdsm
1f9dcea4b6 FN-7935: route mailbox artifact View task to the popped-out task-detail window 
Mailbox artifact "View task" now opens the producing task in the same shared, movable/resizable floating task-detail window used elsewhere in the dashboard, instead of the docked task-detail modal.

- MainContent's MailboxView onOpenTask handler now calls popOutTaskDetail(task) after fetchTaskDetail resolves, instead of openDetailTask(task), matching DocumentsView's artifact-task open path
- add regression test verifying mailbox artifact "View task" clicks resolve the task and route to popOutTaskDetail (not openDetailTask)
- update docs/dashboard-guide.md to describe the shared movable/resizable task-detail window behavior
- add changeset (patch) documenting the fix for @runfusion/fusion

Files changed:
 .changeset/fn-7935-mailbox-artifact-view-task-popout.md                        |  7 ++
 docs/dashboard-guide.md                                                        |  2 +-
 packages/dashboard/app/components/dashboard/MainContent.tsx                    |  8 ++-
 .../MainContent.mailbox-view-task.test.tsx                                     | 83 ++++++++++++++++++++++
 4 files changed, 97 insertions(+), 3 deletions(-)

Fusion-Task-Id: FN-7935

Fusion-Task-Lineage: 51374962-aa36-4390-a6b5-b519e7fc2bf2

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 00:50:30 -07:00
gsxdsm
29560021d3 FN-7934: fix chat brain popup clipping in narrow floating windows 
Fixes the in-chat model/thinking (Brain) popup being cut off inside a narrow floating Chat window or compact dock on a wide desktop viewport.

- Key the popover's viewport-fitting inset layout on ChatView's .chat-view--narrow class (chat surface width) instead of only the @media (max-width: 768px) browser-viewport query, so narrow floating/docked chat surfaces get the fitted layout too.
- Add narrow-surface CSS rules for .chat-thinking-level-root, .chat-thinking-popover, .chat-thinking-agent-list, and .chat-thinking-popover-list to constrain position/width/max-height to the chat surface.
- Add a CSS-contract regression test asserting both the desktop popover sizing and the new narrow-surface rules stay in sync.
- Update docs/dashboard-guide.md to describe the popup staying fitted to the chat surface for narrow floating Chat windows/compact docks, not just mobile/tablet viewports.
- Add a patch changeset for @runfusion/fusion documenting the fix.

Files changed:
 .changeset/fn-7934-chat-narrow-model-popup.md      |  7 +++++
 docs/dashboard-guide.md                            |  3 +-
 packages/dashboard/app/components/ChatView.css     | 22 +++++++++++++++
 .../__tests__/ChatThinkingLevelControl.test.tsx    | 32 ++++++++++++++++++++++
 4 files changed, 63 insertions(+), 1 deletion(-)

Fusion-Task-Id: FN-7934

Fusion-Task-Lineage: 30c461c5-a153-4005-8a8c-24f02916a934

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 00:48:02 -07:00
gsxdsm
9ba8a2e575 FN-7932: add per-lane Reviewer and Planning thinking-level overrides 
Adds validatorThinkingLevel and planningThinkingLevel task fields so the Reviewer and Planning AI lanes can override reasoning effort independently of the shared task thinkingLevel, with dashboard UI, storage, and runtime fallback wiring.

- Add validatorThinkingLevel and planningThinkingLevel to Task/TaskCreateInput types (packages/core/src/types.ts)
- Persist the new fields in the SQLite schema and store read/write/replication paths (packages/core/src/db.ts, store.ts, mesh-task-replication.ts)
- Wire executor and triage lanes to fall back per-lane thinking level -> task.thinkingLevel -> existing settings/lane fallback (packages/engine/src/executor.ts, triage.ts)
- Add per-lane thinking-level selectors to the ModelSelectorTab UI, alongside the existing thinking-level control (packages/dashboard/app/components/ModelSelectorTab.tsx)
- Expose the new fields through the legacy task API and task-workflow routes (packages/dashboard/app/api/legacy.ts, packages/dashboard/src/routes/register-task-workflow-routes.ts)
- Document the new settings in dashboard-guide.md and settings-reference.md
- Add a minor changeset and unit/integration test coverage for store persistence, routes, UI, and agent-session helpers

Files changed:
 .changeset/per-lane-task-thinking.md               |   7 ++
 docs/dashboard-guide.md                            |   2 +
 docs/settings-reference.md                         |   2 +-
 .../src/__tests__/store-thinking-levels.test.ts    |  43 +++++++
 packages/core/src/db.ts                            |  15 ++-
 packages/core/src/mesh-task-replication.ts         |   4 +
 packages/core/src/store.ts                         |  24 +++-
 packages/core/src/types.ts                         |  12 ++
 packages/dashboard/app/api/legacy.ts               |   2 +
 .../dashboard/app/components/ModelSelectorTab.tsx  | 126 ++++++++++++++++++++-
 .../components/__tests__/ModelSelectorTab.test.tsx |  50 +++++++-
 .../src/__tests__/routes-tasks-ops.test.ts         |  74 ++++++++++++
 .../src/routes/register-task-workflow-routes.ts    |  19 +++-
 .../src/__tests__/agent-session-helpers.test.ts    |  15 +++
 packages/engine/src/executor.ts                    |  16 ++-
 packages/engine/src/triage.ts                      |   8 +-
 16 files changed, 395 insertions(+), 24 deletions(-)

Fusion-Task-Id: FN-7932

Fusion-Task-Lineage: 4202f774-aab9-41d2-86a0-f5277dd0f848

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 00:44:59 -07:00
gsxdsm
5f432971db FN-7931: fix terminal workspace picker rendering behind floating terminal modal 
Fixes the portaled terminal workspace picker menu appearing invisible behind the floating terminal modal by z-layering it above the panel's floatingZ and hiding it until it is positioned.

- Compute terminalWorkspaceMenuFloatingZ one layer above the floating modal's floatingZ (min 5000) so the portaled listbox always renders above the floating terminal stack.
- Position the workspace picker menu synchronously via useLayoutEffect before paint, instead of relying only on the async rAF-driven position update.
- Keep the menu invisible and non-interactive (visibility: hidden, pointer-events: none) until computed trigger-relative coordinates are applied, avoiding a flash at stale/fallback CSS coordinates.
- Add regression tests covering floating-mode z-index layering/pre-rAF positioning and docked/below/embedded/mobile workspace-picker positioning.
- Add a patch changeset describing the fix.

Files changed:
 .changeset/terminal-workspace-picker-floating.md   |   7 ++
 .../dashboard/app/components/TerminalModal.tsx     |  37 +++++--
 .../components/__tests__/TerminalModal.test.tsx    | 120 +++++++++++++++++++++
 3 files changed, 156 insertions(+), 8 deletions(-)

Fusion-Task-Id: FN-7931

Fusion-Task-Lineage: 41736804-3f99-4cb7-a0fb-f755a86ffc62

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 00:22:41 -07:00
gsxdsm
27110ed893 FN-7930: remove terminal footer 'Connected' label and shortcut help text 
Declutters the terminal footer by dropping steady-state noise: the redundant 'Connected' status text (the header status dot already conveys connection state) and the persistent zoom/shortcuts/escape help copy.

- Remove .terminal-shortcuts / .terminal-shortcuts--header CSS rules (including responsive breakpoint overrides) as orphaned styling
- Stop rendering the 'Connected' text and the Ctrl++/- zoom / Shortcuts panel / Esc close help span in TerminalModal's footer
- Drop the now-unused terminal.helpText locale key from all 6 locales (en, es, fr, ko, zh-CN, zh-TW) and regenerate resources.d.ts
- Update TerminalModal tests to match the trimmed footer markup and add a regression test asserting the connected-status text and shortcut help are omitted
- Add a patch changeset documenting the fix

Files changed:
 .changeset/remove-terminal-footer-noise.md         |  7 +++++
 .../dashboard/app/components/TerminalModal.css     | 34 ---------------------
 .../dashboard/app/components/TerminalModal.tsx     |  7 +++--
 .../components/__tests__/TerminalModal.test.tsx    | 35 +++++++++++++++++-----
 packages/i18n/locales/en/app.json                  |  1 -
 packages/i18n/locales/es/app.json                  |  1 -
 packages/i18n/locales/fr/app.json                  |  1 -
 packages/i18n/locales/ko/app.json                  |  1 -
 packages/i18n/locales/zh-CN/app.json               |  1 -
 packages/i18n/locales/zh-TW/app.json               |  1 -
 packages/i18n/src/resources.d.ts                   |  1 -
 11 files changed, 40 insertions(+), 50 deletions(-)

Fusion-Task-Id: FN-7930

Fusion-Task-Lineage: 2b8acb5d-c4d4-4766-ba84-b2691a7eb5e8

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 00:07:21 -07:00
gsxdsm
da919fddea FN-7928: align TaskCard header-actions controls on a shared centerline 
Fixes vertical baseline drift among the Send-back button, ⋯ menu button, and size badge inside .card-header-actions by normalizing their line-height, plus adds regression coverage.

- Set line-height: 1 on .card-menu-btn and .card-send-back-btn so their inline-flex content no longer drifts off-center relative to the size badge
- Add FNXC:TaskCardLayout comment documenting the FN-7928 requirement and its coexistence with prior FN-7889/FN-7862/FN-7837/FN-4351 header rules
- Add a badge-wrap regression test asserting Send-back/menu/size controls share one optical centerline across in-progress, done, triage, no-menu, and no-size card states

Files changed:
 packages/dashboard/app/components/TaskCard.css                          |   6 +
 packages/dashboard/app/components/__tests__/TaskCard.badge-wrap.test.tsx | 136 +++++++++++++++++++++
 2 files changed, 142 insertions(+)

Fusion-Task-Id: FN-7928

Fusion-Task-Lineage: 1b4ad644-8495-4ea1-8687-0849d6a3319b

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-13 00:01:42 -07:00
gsxdsm
967f3dd900 FN-7923: align task-card cost badge bottom-right with other footer chips 
Narrative: Reworked TaskCard footer/meta layout so the cost badge (and its sibling footer-right chips) render inline at the bottom-right of the card-meta row when the footer has no leading content, instead of always sitting in a separate footer row beside the time badge.

- Extracted the footer-right chip cluster (cost, time, retry, near-duplicate, undo-of, GitHub tracking) into a shared `footerRightCluster` render, computed once instead of duplicated inline.
- Added `footerHasLeadingContent`/`footerRightHasContent`/`placeFooterRightInMeta` derivations so the cluster moves into `.card-meta` (bottom-right, inline with other tags) when there's no files-changed button or GitHub-import leading content, and the meta row is visible; otherwise it keeps the existing `.card-footer-row` placement for in-progress/tracked cards.
- Updated dashboard-guide.md wording to describe the cost badge as appearing 'with the card's other footer/meta chips' rather than 'beside the execution-time badge'.
- Extended TaskCard.test.tsx coverage for the new placement behavior.
- Desktop local-runtime.ts: kept the previously-unused `reason` parameter on `requestRestart` explicitly referenced (void reason) for API parity/lint cleanliness, unrelated cosmetic cleanup carried in the same branch.

Files changed:
 docs/dashboard-guide.md                            |   2 +-
 packages/dashboard/app/components/TaskCard.tsx     | 242 +++++++++++----------
 .../app/components/__tests__/TaskCard.test.tsx     |  96 +++++++-
 packages/desktop/src/local-runtime.ts              |   4 +-
 4 files changed, 222 insertions(+), 122 deletions(-)

Fusion-Task-Id: FN-7923

Fusion-Task-Lineage: 7e4c3109-f39e-45c0-af13-358ce54f945c

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 23:56:13 -07:00
gsxdsm
30d2e3660d FN-7927: fix Refine feedback modal self-dismissing immediately after opening 
The task-detail Refine overlay used a raw onClick backdrop handler with a stopPropagation-wrapped inner modal, so the same click/touch sequence that opened Refine could bubble into the backdrop handler and close it right away; route it through the shared useOverlayDismiss contract instead so it behaves like every other dashboard modal.

- Compute refineOverlayDismissProps via useOverlayDismiss(handleCloseRefineModal) and spread it onto the refine overlay instead of a plain onClick handler
- Drop the redundant stopPropagation-only onClick from the inner .detail-refine-modal div now that the overlay itself no longer misfires on the opening interaction
- Update docs/dashboard-guide.md to document that the Refine modal (Board and List entry points) stays open until an explicit close or an enabled backdrop dismissal
- Add TaskDetailModal.refine.test.tsx regression coverage for the modal staying open across the opening interaction and honoring the dismiss-preference gate
- Add a patch changeset summarizing the fix for @runfusion/fusion release notes

Files changed:
 .changeset/fn-7927-refine-modal.md                 |   7 +
 docs/dashboard-guide.md                            |   8 +-
 .../dashboard/app/components/TaskDetailModal.tsx   |  12 +-
 .../__tests__/TaskDetailModal.refine.test.tsx      | 182 +++++++++++++++++++++
 4 files changed, 201 insertions(+), 8 deletions(-)

Fusion-Task-Id: FN-7927

Fusion-Task-Lineage: c7e7cd4a-d103-47e6-93ce-6577147b4795

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 23:37:54 -07:00
gsxdsm
f7e942e6f4 fix: resolve all full-suite failures + add structural mock-completeness gate check (round 10) (#2040) 
## Summary

Fixes ALL failing shards from the latest full-suite run (29225946428)
AND adds a structural gate check to prevent the recurring mock-export
drift pattern that has caused every full-suite failure across rounds
1–9.

## What broke (run 29225946428, commit 504b0f8b0)

| Shard | Root cause | Tests fixed |
|---|---|---|
| **3 (CLI)** | `workflowValidateParams` (FN-7911) missing from
`@fusion/engine` mock | 8 files |
| **3 (CLI)** | `skill-sync.test.ts` — `fn_workflow_validate` missing
from engine-tools.md | 1 file |
| **4 (dashboard)** | 6 chat default settings keys missing from
description allowlist | 1 file |
| **1+2 (engine)** | `additionalSkillPaths` missing from
`buildSessionSkillContext` mocks (FN-1510/1511) | 10 tests |
| **1+2 (engine)** | heartbeat FN-7878 changed paused→error for generic
run failures | 1 test |
| **1+2 (engine)** | executor `updateTask` exact-match →
`objectContaining` (new fields) | 2 tests |
| **1+2 (engine)** | `connectMcpSessionTools` mock missing for pi.test
MCP forwarding | 1 test |

## Structural fix — `scripts/check-mock-completeness.mjs` (the "fix for
good")

**New gate check** added to `pnpm test:gate`. Statically validates every
hardcoded `vi.mock("@fusion/dashboard")` and `vi.mock("@fusion/engine")`
factory covers all named imports the source file uses. Runs in <0.2s, no
module evaluation.

**How it works:**
1. Extracts named exports from each barrel
(`packages/dashboard/src/index.ts`, `packages/engine/src/index.ts`)
2. For each test file with a hardcoded `vi.mock` factory (no
`importOriginal`/`importActual` spread):
- Resolves source files the test covers (static + dynamic imports,
convention mapping)
   - Extracts what those source files named-import from the barrel
- Resolves spread helpers (e.g. `...workflowAuthoringEngineMock`) by
reading the helper's exported keys
- Reports any barrel exports that are named-imported by source but
absent from the mock

**Why this fixes the recurring pattern:** Every round 1–9 failure was a
new barrel export imported by source but missing from a test mock. This
check catches it at gate time, before merge — not after the full-suite
fails on main.

Also completed all 15 latent mock gaps the guard found on first run (9
dashboard + 6 engine), including expanding the centralized
`workflowAuthoringEngineMock` helper with all `extension.ts` named
imports.

## Verification
- Gate (with new check): exit 0 ✅
- CLI: 355/355 passed ✅
- Engine (6 fixed files): 250/250 passed ✅
- i18n + settings: verified ✅
- Mock completeness guard: ✅ (0 issues)

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit

* **New Features**
* Documented a new non-destructive workflow validation tool that
performs a dry-run and returns typed validation errors.

* **Tests**
* Updated and strengthened CLI, dashboard, extension, and engine tests
with more accurate mock exports and more resilient assertions.
* Adjusted expectations for session/heartbeat and retry-related
behaviors.

* **Chores**
* Added an automated mock-completeness gate and integrated it into the
test quality gate to keep mocks aligned with available platform exports.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
 2026-07-12 23:29:06 -07:00
gsxdsm
6dcecb0c34 FN-7926: park completed-but-blocked tasks instead of looping execute-requeue 
Stops the execute → pause-abort → re-queue-to-todo infinite loop for tasks whose implementation work is done but a dependency/blockedBy blocker is still live, by diverting them into a dedicated parked state instead of feeding the FN-7863 no-progress backstop or looping forever.

- Add TaskExecutor.parkCompletedBlockedTask(): when work is complete but getTaskCompletionBlocker() still reports a blocker, park the task in todo with pausedReason:"completed-work-blocked", status:"queued", preserved worktree/branch/steps, and a cleared execute-requeue signature.
- Replace shouldFinalizeCompletedTask's boolean with getCompletedTaskFinalizationDecision() returning "finalize" | "blocked" | "incomplete" so both the paused-after-completion and finalization call sites can react to the new "blocked" outcome without re-entering execution.
- Divert completed-but-blocked tasks before the FN-7863 execute-requeue-loop counter increments, so waiting-on-dependency states are no longer misclassified as EXECUTION_DISPATCH_LOOP_EXHAUSTED.
- Add SelfHealingManager.reconcileCompletedBlockedTasks(): a bounded sweep (wired into both startup/maintenance and periodic self-healing passes) that clears the park and advances the task to review once getTaskCompletionBlockerForStore() resolves, guarded by auto-merge eligibility, user-pause, and live-execution checks; failed advances re-park rather than strand the row.
- Add run-audit mutation types task:completed-blocked-parked and task:completed-blocked-advanced (ids/counts/outcomes-only metadata) plus AGENTS.md/docs/architecture.md entries documenting the new lifecycle.
- Extend execute-requeue-loop-guard.test.ts with coverage for the park/advance flow, including the zero-step task edge case.

Files changed:
 AGENTS.md                                          |   1 +
 docs/architecture.md                               |   2 +
 .../execute-requeue-loop-guard.test.ts             | 256 ++++++++++++++++++++-
 packages/engine/src/executor.ts                    |  85 ++++++-
 packages/engine/src/run-audit.ts                   |   4 +
 packages/engine/src/self-healing.ts                |  95 ++++++++
 6 files changed, 432 insertions(+), 11 deletions(-)

Fusion-Task-Id: FN-7926
Fusion-Task-Lineage: e47945f4-a816-447e-9ea1-7c13105d0ba9
Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 23:26:02 -07:00
gsxdsm
e84fda936a FN-7918: make chat go-to-top contextual and inline edit pencil compact 
Reworks chat message footer affordances: the scroll-to-top control now only becomes visible once a message's top is actually clipped above the visible thread viewport, and the edit pencil moves from a standalone action row into the timestamp footer beside user messages.

- ChatView measures assistant message tops on scroll/message changes (rAF-scheduled) and tracks which message IDs are currently clipped above the `.chat-messages` container edge
- StandardChatMessageItem accepts a new `isTopClipped` prop; the go-to-top button stays DOM-mounted (for tests/a11y) but is visually hidden via CSS until clipped
- Merged the assistant thinking/copy/scroll-to-top actions into a single collapsible footer row instead of separate action rows
- Moved the user-message edit pencil into an inline `chat-message-time-row` next to the relative timestamp instead of a standalone action row above it
- Updated ChatView.css for the new inline layout, collapsed-row state, and hidden/visible scroll-to-top button states
- Updated message-edit and scroll-to-top tests to cover the new inline placement and clipped-visibility behavior
- Added changeset and docs/dashboard-guide.md note describing the new behavior

Files changed:
 .changeset/fn-7918-chat-inline-icons.md            |  7 ++
 docs/dashboard-guide.md                            |  6 +-
 packages/dashboard/app/components/ChatView.css     | 80 +++++++++++++++-------
 packages/dashboard/app/components/ChatView.tsx     | 52 +++++++++++++-
 .../app/components/StandardChatSurface.tsx         | 33 +++++++--
 .../__tests__/ChatView.message-edit.test.tsx       | 34 ++++++++-
 .../__tests__/ChatView.scroll-to-top.test.tsx      | 75 +++++++++++++++++++-
 7 files changed, 253 insertions(+), 34 deletions(-)

Fusion-Task-Id: FN-7918
Fusion-Task-Lineage: 76206cd2-94a8-47be-b282-94943e184d01
Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 23:24:03 -07:00
gsxdsm
c8999369c3 feat: add gate check for CLI dashboard mock completeness — prevents recurring full-suite barrel-export drift (#2035) 
## Summary

**Structural fix** for the recurring full-suite failure pattern where a
new `@fusion/dashboard` barrel export is imported by CLI source code but
missing from the hardcoded `vi.mock("@fusion/dashboard")` factory in CLI
tests.

## What's new

### Gate check script:
`scripts/check-cli-dashboard-mock-completeness.mjs`
Added to the merge gate (`pnpm test:gate`). Statically validates that
every hardcoded `vi.mock("@fusion/dashboard")` factory in CLI tests
includes all `@fusion/dashboard` exports that the corresponding source
files import.

- Pure static analysis (regex + depth-aware brace tracking) — no module
evaluation, <0.1s
- Handles named imports (`import { foo } from "@fusion/dashboard"`) AND
namespace imports (`import * as dashboard from "@fusion/dashboard"` →
scans `dashboard.X` usages)
- Filters against the real barrel exports to avoid false positives from
typos
- Resolves test→source mapping by parsing static/dynamic imports in the
test file (not just naming convention)

**Result:** the next time someone adds `export { newFunc } from
"./mod.js"` to `dashboard/src/index.ts` and `cli/src/commands/daemon.ts`
imports it, the gate catches the missing mock before merge instead of
the full-suite failing on main.

### Completed all 9 incomplete CLI dashboard mocks
Added the missing exports identified by the check:

| File | Missing exports added |
|---|---|
| `daemon.test.ts` | `registerGithubTrackingHook` |
| `serve.test.ts` | `registerGithubTrackingHook` |
| `dashboard.test.ts` | `AttachTicketStore`, `CliInputAttributionLog`,
`CliConfirmAdvanceRegistry`, `CliRelaunchRegistry`,
`registerGithubTrackingHook` |
| `task.test.ts` | `registerGithubTrackingHook`, `GitLabClient`,
`resolveGitlabAuth`, `buildGitLabTaskProvenance`,
`isGitLabAlreadyImported`, `buildGitLabTaskDescription` |
| `extension-*.test.ts` (×4) | `GitLabClient`, `resolveGitlabAuth`,
`buildGitLabTaskProvenance`, `isGitLabAlreadyImported`,
`buildGitLabTaskDescription` |
| `task-command-github-import-tracking.test.ts` | Same GitLab exports |

These were latent issues — the mocks were incomplete but tests passed
because the missing exports weren't called during test execution. Any
test change that exercises those code paths would have broken.

## Why not `importActual` spread?
Tried converting daemon.test.ts to `vi.mock("@fusion/dashboard", async
(importOriginal) => { ... })` — fails because the barrel's `export *
from "./plugins/index.js"` transitively imports
`@agentclientprotocol/sdk` which isn't available at test evaluation
time. The static check approach avoids this entirely.

## Verification
- `pnpm test:gate`: exit 0 (includes new check)
- `pnpm lint`: exit 0
- CLI tests: daemon 21/21, serve 58/58, dashboard 91/91, task 149/149 ✅
- Gate script: `✅ CLI dashboard mock completeness: all hardcoded mocks
cover source imports.`

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

- **Tests**
- Added automated validation to ensure CLI test mocks remain aligned
with available dashboard functionality.
- Updated test coverage setup so GitHub, GitLab, daemon, dashboard,
server, and task scenarios use complete dashboard mocks.
- Test verification now reports missing mocked functionality and blocks
the release gate when inconsistencies are detected.

- **Chores**
- Improved reliability and maintainability of automated verification for
CLI and dashboard integrations.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
 2026-07-12 23:09:50 -07:00
gsxdsm
2aefaad319 FN-7924: Add View task link to artifact-registration mail notifications 
Artifact-registration mail messages now expose the task that produced the artifact so users can jump straight to it from the mailbox.

- MailboxArtifactAttachment renders a "View task" button when message.metadata.taskId is present and an onOpenTask handler is supplied, alongside the existing Open artifact affordance
- MailboxModal and MailboxView thread taskId metadata and onOpenTask through to MailboxArtifactAttachment for both the message-list and detail-pane renders
- MainContent wires MailboxView's onOpenTask to the shared fetchTaskDetail -> openDetailTask path, with a toast on failure, so mailbox reuses the existing task-detail flow
- docs/dashboard-guide.md documents the new View task affordance for artifact notifications
- adds a minor changeset for @runfusion/fusion describing the new mail notification behavior
- extends MailboxArtifactAttachment and MailboxView tests to cover the new taskId/onOpenTask wiring

Files changed:
 .changeset/fn-7924-artifact-mail-view-task-link.md |  7 ++++
 docs/dashboard-guide.md                            |  2 +-
 .../app/components/MailboxArtifactAttachment.tsx   | 20 ++++++++++
 packages/dashboard/app/components/MailboxModal.css |  2 +-
 packages/dashboard/app/components/MailboxModal.tsx |  6 +++
 packages/dashboard/app/components/MailboxView.tsx  |  6 +++
 .../__tests__/MailboxArtifactAttachment.test.tsx   | 35 ++++++++++++++++-
 .../app/components/__tests__/MailboxView.test.tsx  | 45 +++++++++++++++++++++-
 .../app/components/dashboard/MainContent.tsx       |  6 +++
 9 files changed, 124 insertions(+), 5 deletions(-)

Fusion-Task-Id: FN-7924

Fusion-Task-Lineage: 800102e1-9025-40da-8130-9ab0e8acd747

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 23:06:05 -07:00
gsxdsm
382a4d5d05 FN-7917: center chat composer attach/model icons with input box 
Centers the Chat composer's attach and model/thinking icon buttons with the single-line message input across desktop, tablet, and mobile, while keeping them bottom-aligned when the textarea grows multi-line.

- Size .chat-attach-btn, .chat-thinking-level-root, and .chat-thinking-btn to --chat-input-control-size (matching the send button/textarea min-height) instead of the old calc(var(--space-lg) * 2) / * 2.25) sizing
- Replace the separate mobile touch-target overrides for .chat-attach-btn and .chat-thinking-btn with a shared rule using the same --chat-input-control-size sizing
- Add a regression test asserting the attach/thinking controls use --chat-input-control-size and align-self: flex-end, and no longer use the old calc()-based sizing
- Add a patch changeset for @runfusion/fusion documenting the composer icon centering fix

Files changed:
 .changeset/chat-composer-icon-centering.md         |  7 +++++
 packages/dashboard/app/components/ChatView.css     | 30 ++++++++++++----------
 .../ChatView.chat-input-autosize.test.tsx          | 28 ++++++++++++++++++++
 3 files changed, 52 insertions(+), 13 deletions(-)

Fusion-Task-Id: FN-7917
Fusion-Task-Lineage: aba1d5ea-21ee-4b27-90f2-11aca7c4a5fa
Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 23:02:46 -07:00
gsxdsm
87aab438dc FN-7922: fix jerky tablet drag on headerless floating-window handles 
Ensures every draggable modal surface (terminal, artifacts viewer, and other
FloatingWindow delegates) keeps touch-action: none on its drag handle so
tablet touch-drag is captured by FloatingWindow's pointermove stream instead
of being intersected by page pan, and backs the contract with tests.

- Add touch-action: none to .artifacts-gallery-viewer-header, the headerless
  FloatingWindow's delegated drag handle, matching the other movable modal
  handles.
- Add a FloatingWindow test asserting touch-action: none is present across
  every tablet movable-modal drag handle (terminal, artifacts, workflow
  editor, automation, mission interview, PR create, file browser, right
  dock, new task modal, quick chat FAB) and absent from broad tablet
  pan-y overrides.
- Add a FloatingWindow test exercising the captured tablet touch-drag path
  for a headerless delegated handle (pointerdown/move/up with
  pointerType: "touch").
- Fix TerminalModal drag test to send pointerType: "touch" on
  pointerdown/move/up so it exercises the same tablet touch-drag contract.
- Add changeset for the tablet drag fix.

Files changed:
 .changeset/fn-7922-tablet-modal-drag.md            |  7 ++
 .../dashboard/app/components/ArtifactsGallery.css  |  5 ++
 .../components/__tests__/FloatingWindow.test.tsx   | 76 +++++++++++++++++++++-
 .../components/__tests__/TerminalModal.test.tsx    |  6 +-
 4 files changed, 90 insertions(+), 4 deletions(-)

Fusion-Task-Id: FN-7922

Fusion-Task-Lineage: 0d63ced5-faa7-4362-a42d-884b4c889ebf

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 22:59:52 -07:00
gsxdsm
313956df5a FN-7916: fix chat model selector dismissing before selection on mobile/tablet 
Fix the in-chat Brain popup's model picker on touch viewports: pointerdown outside-close was treating the portaled CustomModelDropdown menu as outside the popup, dismissing it before the tap could register a model selection, and the mobile popover was mis-anchored off-screen.

- ChatThinkingLevelControl: treat pointerdown targets inside the portaled `.model-combobox-dropdown--portal` menu as inside the popup so touch taps select the model instead of closing the popup first
- ChatView.css: clamp the popover to the viewport width via max-width/max-inline-size, and on mobile anchor it to the chat input area with tokenized left/right gutters instead of a collapsing left:0 width
- Add a portal-aware regression test covering pointerdown-inside-portal selection, genuine outside-pointerdown close, and the mobile CSS anchoring contract
- Add changeset (patch) and update dashboard guide docs to describe mobile/tablet behavior

Files changed:
 .changeset/fn-7916-chat-mobile-model-selector.md   |  7 ++
 docs/dashboard-guide.md                            |  3 +-
 .../app/components/ChatThinkingLevelControl.tsx    | 10 ++-
 packages/dashboard/app/components/ChatView.css     | 17 +++-
 .../ChatThinkingLevelControl.portal.test.tsx       | 97 ++++++++++++++++++++++
 5 files changed, 130 insertions(+), 4 deletions(-)

Fusion-Task-Id: FN-7916

Fusion-Task-Lineage: bcc2d9ad-0be6-4e23-b643-eebf93aae3c6

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 22:49:51 -07:00
gsxdsm
d2f8acd4f1 FN-7920: remove duplicate dollar-sign icon from task-card cost badges 
Removes the redundant DollarSign icon rendered inside the task-card cost badge, since the formatted cost label already includes a currency symbol and showing both duplicated the "$".

- Drop the leading DollarSign icon from the cost badge span in TaskCard.tsx (and its now-unused lucide-react import)
- Add an FNXC comment documenting that formatCost already includes the currency symbol, so no icon should duplicate it
- Update TaskCard tests to assert the cost badge renders no icon (svg) and to cover aria-label/title text for both populated and empty-cost states

Files changed:
 packages/dashboard/app/components/TaskCard.tsx                | 7 +++++--
 packages/dashboard/app/components/__tests__/TaskCard.test.tsx | 5 ++++-
 2 files changed, 9 insertions(+), 3 deletions(-)

Fusion-Task-Id: FN-7920

Fusion-Task-Lineage: 9404ba50-944b-430d-b7ad-0613ec75c42d

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 22:47:07 -07:00
gsxdsm
504b0f8b04 FN-7914: fix mobile prompt fullscreen editor Expand button hidden behind workflow sheet 
Fixes the workflow editor's fullscreen prompt overlay having an inert-looking mobile Expand button because a static z-index of 10000 sat below the mobile workflow editor's full-screen FloatingWindow sheet (10100+ shared floating-stack band).

- Have the fullscreen prompt overlay claim a fresh shared floating-stack z-index via nextFloatingZ() when opened, instead of relying on the static CSS z-index: 10000 fallback
- Track the claimed z-index in new promptFullscreenZ state, applying it as an inline style on the portaled overlay and clearing it on collapse/Escape
- Update FNXC comments in WorkflowNodeEditor.css and floatingWindowStack.ts documenting the shared z-index contract and why the static value is now only a fallback
- Add regression tests asserting the prompt fullscreen overlay's z-index exceeds the workflow editor's floating window z-index on desktop, mobile prompt nodes, mobile gate nodes, and after re-opening after collapse

Files changed:
 .../app/components/WorkflowNodeEditor.css          |  1 +
 .../app/components/WorkflowNodeEditor.tsx          | 25 ++++++-
 .../__tests__/WorkflowNodeEditor.test.tsx          | 81 ++++++++++++++++++++++
 .../app/components/floatingWindowStack.ts          |  2 +-
 4 files changed, 106 insertions(+), 3 deletions(-)

Fusion-Task-Id: FN-7914

Fusion-Task-Lineage: e230309a-3b53-4e42-9e63-7516597b6c24

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 22:15:48 -07:00
gsxdsm
daf3f15fd1 FN-7909: add room-level thinking effort override for Chat Rooms 
Adds a per-room thinking-effort (reasoning level) override for Chat Rooms so all room responders can share a consistent override instead of relying only on per-agent/global defaults.

- Persist `chat_rooms.thinkingLevel` with a new core DB migration and store read/write support
- Extend chat-store and chat-types with thinkingLevel plumbing for room create/update
- Wire the dashboard chat room API/routes and legacy handlers to accept and return thinkingLevel
- Add a ChatView room settings control (with CSS) and useChatRooms hook support for setting/clearing the override
- Resolve room responder defaultThinkingLevel from the room override when present
- Update docs (dashboard-guide, settings-reference) and add a minor changeset for the feature

Files changed:
 .changeset/fn-7909-room-thinking-level.md          |  7 +++
 docs/dashboard-guide.md                            |  1 +
 docs/settings-reference.md                         |  2 +-
 packages/core/src/__tests__/chat-store.test.ts     | 21 ++++++++
 packages/core/src/__tests__/db-migrate.test.ts     | 57 ++++++++++++++++++++++
 packages/core/src/chat-store.ts                    | 12 ++++-
 packages/core/src/chat-types.ts                    | 10 ++++
 packages/core/src/db.ts                            | 28 ++++++++++-
 packages/dashboard/app/api/__tests__/chat-rooms-api.test.ts       |  8 +--
 packages/dashboard/app/api/legacy.ts               |  4 +-
 packages/dashboard/app/components/ChatView.css     | 16 ++++++
 packages/dashboard/app/components/ChatView.tsx     | 29 ++++++++++-
 packages/dashboard/app/components/__tests__/ChatView.rooms.test.tsx   | 24 +++++++++
 packages/dashboard/app/hooks/__tests__/useChatRooms.test.ts       | 22 +++++++++
 packages/dashboard/app/hooks/useChatRooms.ts       | 16 ++++++
 packages/dashboard/src/__tests__/chat-room-routes.test.ts         | 26 ++++++++++
 packages/dashboard/src/__tests__/chat.rooms.test.ts     | 42 ++++++++++++++++
 packages/dashboard/src/chat.ts                     |  8 +++
 packages/dashboard/src/routes/register-chat-room-routes.ts        | 21 ++++++--
 19 files changed, 338 insertions(+), 16 deletions(-)

Fusion-Task-Id: FN-7909

Fusion-Task-Lineage: 2741eca9-5305-4f6c-81bf-ae644a9fe307

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 22:12:18 -07:00
gsxdsm
4ea5dd6739 fix: CLI shouldSuperviseDashboard mock + getCliPackageVersion/isUnresolvedCliPackageVersion + executeRequeueLoopCount + i18n buttonShort (round 9) (#2034) 
## Summary

Fixes shard 3 (CLI) + shard 4 (i18n) failures from full-suite run
29223788088.

## Fixes

### CLI (shard 3) — 5 files, ~140 tests
- **`bin.test.ts`** — `bin.ts:851` now imports
`shouldSuperviseDashboard` and `bin.ts:865` imports
`runDashboardSupervised` from `./commands/dashboard.js`. Test mock only
exported `runDashboard`. Missing `shouldSuperviseDashboard` → TypeError
→ caught → `process.exit:1`. Added `shouldSuperviseDashboard: vi.fn(()
=> false)` + `runDashboardSupervised` to mock + `commandMocks`.
- **`daemon/serve/dashboard.test.ts`** — `@fusion/dashboard` barrel
(index.ts:115) re-exports `getCliPackageVersion`,
`isUnresolvedCliPackageVersion`, `resolveCliPackageVersionInfo`. Added
all 3 to each file's `@fusion/dashboard` mock.
- **`task.test.ts`** — `executeRequeueLoopCount: 0` added to
TaskResetField set by recent commit; retry test assertions needed the
new field in both `mockUpdateTask` expectations.

### i18n (shard 4) — 4 locale files
- **`settings.reset.buttonShort`** missing from zh-TW, fr, es, ko (zh-CN
already had it).

### Engine (shards 1+2) — already fixed in PR #2025 (merged on main)
- Run 29223788088 was at commit `b85a6b866` (pre-PR-2025-merge), so
engine + i18n `storageMigrationNotice` failures visible in that run are
already resolved on main.

## Verification
- daemon: 21/21 ✅ | serve: 58/58 ✅ | dashboard: 91/91 ✅ | task retry:
5/5 ✅
- i18n parity + gate-coverage: 7/7 ✅
- Gate (`pnpm test:gate`): exit 0 ✅
- `bin.test.ts`: cannot verify locally (`@agentclientprotocol/sdk` not
installed locally; CI resolves from lockfile). Mock exports verified
against `dashboard.ts` source.

<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **Bug Fixes**
* Improved task retry recovery by correctly resetting execution
counters.

* **Localization**
* Added support for shorter reset-button labels in Spanish, French,
Korean, and Traditional Chinese.

* **Tests**
* Updated command and startup checks to reflect current dashboard and
version-handling behavior.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->
 2026-07-12 21:58:12 -07:00
gsxdsm
a628a13888 feat: stamp store-open provenance into run-audit 
Every TaskStore.init() now records a `store:open` run-audit event with
pid/ppid/execPath/entry/cwd/node version. Motivated by the FN-7910 incident:
a stale pre-fix binary opened the shared fusion.db and evacuated Ideas cards,
and the audit trail (agentId:"system", no PID) could not identify the writer.
Any future mystery mutation is now attributable to the process that opened
the store.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
 2026-07-12 21:57:29 -07:00
gsxdsm
1ea185daa5 FN-7911: add workflow validate dry-run command, tool, and API route 
Adds a non-mutating `fn workflow validate` dry-run path across CLI, agent tools, and dashboard API so custom workflow IR can be checked before create/update.

- Add `packages/cli/src/commands/workflow.ts` implementing `fn workflow validate <id> | --file <path>` with JSON/text output, wired into `bin.ts`.
- Add `fn_workflow_validate` agent tool (`agent-tools.ts`, `index.ts`) reusing the existing parseWorkflowIr/trait/code-node/column-agent validation used by create/update, performing no persistence.
- Add `POST /api/workflows/validate` route in `register-workflow-routes.ts` plus dashboard route test coverage.
- Extend heartbeat tool-gating/exposure tests and gating classifications to include `fn_workflow_validate` alongside the other workflow tools.
- Update CLI/agent extension docs (`docs/cli-reference.md`, `docs/agents.md`, `docs/workflow-steps.md`, fusion skill references) to document the new command/tool.
- Add changeset `.changeset/fn-7911-workflow-validate.md` (minor) describing the new capability.

Files changed:
 .changeset/fn-7911-workflow-validate.md            |   7 ++
 docs/agents.md                                     |   5 +-
 docs/cli-reference.md                              |  13 ++
 docs/workflow-steps.md                             |   3 +-
 packages/cli/skill/fusion/SKILL.md                 |   2 +-
 .../cli/skill/fusion/references/extension-tools.md |  10 ++
 .../skill/fusion/references/fusion-capabilities.md |   1 +
 .../src/__tests__/extension-workflow-tools.test.ts |   1 +
 packages/cli/src/__tests__/extension.test.ts       |   1 +
 .../src/__tests__/workflow-docs-current.test.ts    |   1 +
 packages/cli/src/bin.ts                            |  22 ++++
 packages/cli/src/commands/workflow.ts              |  80 ++++++++++++
 packages/cli/src/extension.ts                      |  10 ++
 .../dashboard/src/__tests__/chat-manager.test.ts   |   1 +
 .../dashboard/src/__tests__/chat.rooms.test.ts     |   1 +
 .../planning-document-tools-exposure.test.ts       |   1 +
 .../__tests__/workflow-validate-route.test.ts      | 101 +++++++++++++++
 .../src/routes/register-workflow-routes.ts         |  27 +++-
 .../engine/src/__tests__/agent-action-gate.test.ts |   2 +-
 .../agent-workflow-tools-exposure.test.ts          |  70 ++++++++++-
 .../src/__tests__/gating-classifications.test.ts   |   3 +-
 .../src/__tests__/heartbeat-executor.test.ts       |  37 +++---
 .../src/__tests__/heartbeat-session-prompt.test.ts |   5 +-
 .../src/__tests__/permanent-agent-gating.test.ts   |   2 +-
 packages/engine/src/agent-heartbeat.ts             |   5 +-
 packages/engine/src/agent-tools.ts                 | 140 ++++++++++++++++++++-
 packages/engine/src/executor.ts                    |   6 +
 packages/engine/src/gating-classifications.ts      |   2 +
 packages/engine/src/index.ts                       |   4 +
 29 files changed, 532 insertions(+), 31 deletions(-)

Fusion-Task-Id: FN-7911

Fusion-Task-Lineage: 903d15fe-a7ec-458f-aa34-8f2e895a9603

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 21:39:23 -07:00
gsxdsm
8835c6cb48 FN-7908: add in-chat model/agent switcher to brain-icon popup 
Extend the chat brain-icon popup and its backing session PATCH route so an
active Direct chat's model or agent can be switched mid-conversation instead
of only being set at creation time.

- Add a Model/Agent section to ChatThinkingLevelControl (the brain-icon
  popup) for picking a model provider/model or retargeting to a real agent
  without leaving the chat.
- Extend PATCH /api/chat/sessions/:id to accept modelProvider/modelId (as a
  validated pair via the existing validateModelPair helper) and agentId,
  forwarding only the keys present in the body so omitted fields leave the
  session's stored target untouched.
- Add chat-store updateSession support for the agentId clause alongside the
  existing model/thinkingLevel fields, and a useChat.setSessionModel hook
  for the dashboard to call the new PATCH capability.
- Update i18n locale strings (en/es/fr/ko/zh-CN/zh-TW) and dashboard-guide.md
  docs for the new switcher UI.
- Add unit/integration test coverage across chat-store, chat-manager,
  chat-routes, useChat, ChatThinkingLevelControl, and ChatView for the new
  model/agent switch behavior.
- Add changeset fn-7908-chat-model-agent-switcher.md (minor,
  @runfusion/fusion).

Files changed:
 .changeset/fn-7908-chat-model-agent-switcher.md    |   7 +
 docs/dashboard-guide.md                            |   3 +-
 packages/core/src/__tests__/chat-store.test.ts     |  21 ++
 packages/core/src/chat-store.ts                    |   8 +
 packages/core/src/chat-types.ts                    |   2 +
 packages/dashboard/app/api/legacy.ts               |  11 +-
 .../app/components/ChatThinkingLevelControl.tsx    | 219 ++++++++++++++++++---
 packages/dashboard/app/components/ChatView.css     | 135 ++++++++++++-
 packages/dashboard/app/components/ChatView.tsx     |  23 ++-
 .../__tests__/ChatThinkingLevelControl.test.tsx    | 109 +++++++++-
 .../__tests__/ChatView.thinking-level.test.tsx     |  67 ++++++-
 .../dashboard/app/hooks/__tests__/useChat.test.ts  | 166 +++++++++++++++-
 packages/dashboard/app/hooks/useChat.ts            |  56 ++++++
 .../dashboard/src/__tests__/chat-manager.test.ts   |  38 ++++
 .../dashboard/src/__tests__/chat-routes.test.ts    | 117 ++++++++++-
 .../dashboard/src/routes/register-chat-routes.ts   |  48 ++++-
 packages/i18n/locales/en/app.json                  |   8 +-
 packages/i18n/locales/es/app.json                  |   8 +-
 packages/i18n/locales/fr/app.json                  |   8 +-
 packages/i18n/locales/ko/app.json                  |   8 +-
 packages/i18n/locales/zh-CN/app.json               |   8 +-
 packages/i18n/locales/zh-TW/app.json               |   8 +-
 22 files changed, 1007 insertions(+), 71 deletions(-)

Fusion-Task-Id: FN-7908

Fusion-Task-Lineage: b1104865-9b0c-4d77-973e-89152fe245e0

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 21:33:50 -07:00
gsxdsm
3326984a6d FN-7913: add fn plugin publish --dry-run preflight command 
Adds a non-mutating `fn plugin publish` CLI command that preflights a plugin before manual pack/publish, giving external plugin authors an offline readiness check.

- New `packages/cli/src/commands/plugin-publish.ts` with `runPluginPublish`, `collectPluginPreflight`, and `classifyVersionBump` (strict x.y.z semver bump classification), reusing `loadManifestFromPath` / `resolvePluginEntryFile` from the install path
- Wire `fn plugin publish <path> [--dry-run] [--previous-version <semver>]` into `bin.ts` command routing, dynamic import list, and help text
- Add test coverage in `plugin-publish.test.ts` and update `bin.test.ts` for the new subcommand
- Update `docs/PLUGIN_AUTHORING.md`, `docs/cli-reference.md`, and `docs/plugins/external-authoring.md` to document the new preflight command
- Add changeset `.changeset/fn-7913-plugin-publish-dry-run.md` (minor, @runfusion/fusion)

Files changed:
 .changeset/fn-7913-plugin-publish-dry-run.md      |   7 +
 docs/PLUGIN_AUTHORING.md                          |   9 +-
 docs/cli-reference.md                             |   5 +-
 docs/plugins/external-authoring.md                |  14 +-
 packages/cli/src/__tests__/bin.test.ts            |   2 +-
 packages/cli/src/__tests__/plugin-publish.test.ts | 197 ++++++++++++++++
 packages/cli/src/bin.ts                           |  22 +-
 packages/cli/src/commands/plugin-publish.ts       | 272 ++++++++++++++++++++++
 8 files changed, 521 insertions(+), 7 deletions(-)

Fusion-Task-Id: FN-7913

Fusion-Task-Lineage: 27bdf937-3195-4619-9d01-b6af4fbba487

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 21:27:38 -07:00
gsxdsm
b85a6b8663 FN-7912: add quarantine-ledger deadline visibility check 
Add a report-only script that surfaces flaky-test quarantine entries approaching their 14-day deletion clock, so maintainers can make deliberate rescue-or-expire decisions before entries silently expire.

- Add scripts/check-quarantine-ledger.mjs: reads scripts/lib/test-quarantine.json, computes days-remaining against the existing 14-day deletion clock (shared DELETION_CLOCK_DAYS from scripts/test-velocity-baseline.mjs), and buckets each entry as expired/near/healthy/unknown
- Support --warn-within=<days> (default 5) to tune the near-deadline window, --json for machine-readable output, and --strict as an opt-in local/CI gate (exits 1 on expired/near entries) while default mode stays exit-0 and non-blocking
- Wire pnpm check:quarantine-ledger script in package.json
- Add scripts/__tests__/check-quarantine-ledger.test.mjs covering deadline bucketing/sorting, empty/missing ledger handling, --strict behavior, and --json output shape
- Document the new command and its flags in docs/testing.md under the quarantine ledger/deletion ratchet section

Files changed:
 docs/testing.md                                    |  10 +
 package.json                                       |   1 +
 scripts/__tests__/check-quarantine-ledger.test.mjs | 159 ++++++++++++++++
 scripts/check-quarantine-ledger.mjs                | 202 +++++++++++++++++++++
 4 files changed, 372 insertions(+)

Fusion-Task-Id: FN-7912

Fusion-Task-Lineage: c08e2e09-473a-4ad0-8c27-43cbc3355168

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 21:21:10 -07:00
gsxdsm
03bca1787b FN-7907: Add project chat default (model/agent) setting and New Chat default behavior 
Adds a project-scoped Direct-chat default (model pair or durable agent) plus a New Chat behavior toggle (prompt vs. always-use-default), surfaced in Project Models settings using the standard model dropdown.

- Add ProjectSettings fields: chatNewSessionMode, chatDefaultKind, chatDefaultAgentId, chatDefaultModelProvider, chatDefaultModelId, chatDefaultThinkingLevel
- Extend settings-schema validation for the new chat default fields
- Add a "Chat" subsection to ProjectModelsSection with New Chat behavior selector, model/agent target toggle, and the standard CustomModelDropdown for model selection
- Update ChatView.tsx handleNewChat() flow to honor the configured default (prompt vs. immediate session creation)
- Add i18n strings across en/es/fr/ko/zh-CN/zh-TW locales and regenerate resources.d.ts
- Add changeset (@runfusion/fusion minor) and update docs/settings-reference.md and docs/dashboard-guide.md
- Add settings-parity, ChatView new-chat-default, and ProjectModelsSection chatDefault test coverage

Files changed:
 .changeset/fn-7907-chat-default.md                 |   7 +
 docs/dashboard-guide.md                            |   3 +-
 docs/settings-reference.md                         |   8 +
 packages/core/src/__tests__/settings-parity.test.ts     |  20 ++
 packages/core/src/settings-schema.ts               |   6 +
 packages/core/src/types.ts                         |  15 +
 packages/dashboard/app/components/ChatView.tsx     |  93 +++++-
 packages/dashboard/app/components/__tests__/ChatView.new-chat-default.test.tsx   | 357 +++++++++++++++++++++
 packages/dashboard/app/components/settings/sections/ProjectModelsSection.tsx     |  77 +++++
 packages/dashboard/app/components/settings/sections/__tests__/ProjectModelsSection.chatDefault.test.tsx |  193 +++++++++++
 packages/i18n/locales/en/app.json                  |  20 +-
 packages/i18n/locales/es/app.json                  |  20 +-
 packages/i18n/locales/fr/app.json                  |  20 +-
 packages/i18n/locales/ko/app.json                  |  20 +-
 packages/i18n/locales/zh-CN/app.json               |  20 +-
 packages/i18n/locales/zh-TW/app.json               |  20 +-
 packages/i18n/src/resources.d.ts                   |  18 ++
 17 files changed, 899 insertions(+), 18 deletions(-)

Fusion-Task-Id: FN-7907

Fusion-Task-Lineage: 86a81b9c-7e0c-4033-9e2a-6f7380810fe9

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 20:54:10 -07:00
gsxdsm
41168e273c FN-7905: surface resolved default in chat thinking-level labels 
Chat thinking-level Default entries hardcoded "off" instead of showing the project/global resolved default, misleading operators about which level would actually apply.

- ChatView now fetches Settings via fetchSettings on mount and derives resolvedDefaultThinkingLevel from settings.defaultThinkingLevel (falling back to "off")
- Threads resolvedDefaultThinkingLevel into NewChatDialog and the in-chat ChatThinkingLevelControl (Brain popup) so both surfaces show e.g. "Default (medium)" instead of always "Default (off)"
- ChatThinkingLevelControl accepts a new optional defaultThinkingLevel prop (defaults to "off" for backward compatibility) used only for the Default/clear option label
- Updates dashboard-guide.md to document the resolved-default label behavior in both the New Chat dialog and the in-chat Brain popup
- Adds/updates tests: new ChatThinkingLevelControl coverage for the label prop, new ChatView.thinking-level test coverage for the resolved default, and mocks fetchSettings in the other ChatView test suites that render ChatView
- Adds a patch changeset for @runfusion/fusion

Files changed:
 .changeset/fn-7905-chat-thinking-default.md        |  7 ++++
 docs/dashboard-guide.md                            |  5 +--
 .../app/components/ChatThinkingLevelControl.tsx    |  9 +++--
 packages/dashboard/app/components/ChatView.tsx     | 36 ++++++++++++++++---
 .../__tests__/ChatThinkingLevelControl.test.tsx    | 16 +++++++++
 .../__tests__/ChatView.chat-commands.test.tsx      |  1 +
 .../__tests__/ChatView.content-search.test.tsx     |  1 +
 .../__tests__/ChatView.context-window.test.tsx     |  1 +
 .../__tests__/ChatView.copy-response.test.tsx      |  1 +
 .../__tests__/ChatView.core-contracts.test.tsx     |  1 +
 .../__tests__/ChatView.core-interactions.test.tsx  |  1 +
 .../components/__tests__/ChatView.core.test.tsx    |  1 +
 .../__tests__/ChatView.default-model-icon.test.tsx |  1 +
 .../__tests__/ChatView.hash-mention.test.tsx       |  1 +
 .../components/__tests__/ChatView.mobile.test.tsx  |  1 +
 .../__tests__/ChatView.sessions-rooms.test.tsx     |  1 +
 .../__tests__/ChatView.streaming-thread.test.tsx   |  1 +
 .../__tests__/ChatView.thinking-level.test.tsx     | 42 ++++++++++++++++++++++
 18 files changed, 119 insertions(+), 8 deletions(-)

Fusion-Task-Id: FN-7905

Fusion-Task-Lineage: 7290aa8a-f812-43c3-8ba5-fc31a3f4579e

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 20:19:36 -07:00
gsxdsm
9f8db7d1b5 FN-7903: wire thinkingLevel into AI session creation for automation steps 
Threads the persisted per-step Thinking Level (from FN-7900) into runtime AI session creation and task spawning, so scheduled, routine, and manual automation runs actually apply the chosen reasoning effort instead of only storing it.

- CronRunner passes step.thinkingLevel through AiPromptExecutor to createFnAgent's defaultThinkingLevel for scheduled AI-prompt steps
- RoutineRunner forwards step.thinkingLevel to the shared AiPromptExecutor seam for routine AI-prompt steps
- Cron/routine create-task steps map step.thinkingLevel onto TaskCreateInput.thinkingLevel so spawned tasks inherit the configured reasoning effort
- Dashboard's inline/manual AI-prompt and create-task automation routes apply the same defaultThinkingLevel / TaskCreateInput.thinkingLevel behavior
- Updated docs (dashboard-guide.md, settings-reference.md) to describe the now-active runtime behavior
- Added a changeset for @runfusion/fusion (minor) and expanded cron-runner/routine-runner/routes-automation test coverage

Files changed:
 .changeset/fn-7903-automation-thinking-level.md    |  7 ++
 docs/dashboard-guide.md                            |  5 +-
 docs/settings-reference.md                         |  2 +-
 .../src/__tests__/routes-automation.test.ts        | 67 +++++++++++++++++++
 packages/dashboard/src/routes.ts                   | 10 +++
 packages/engine/src/__tests__/cron-runner.test.ts  | 75 +++++++++++++++++++++-
 .../engine/src/__tests__/routine-runner.test.ts    | 68 +++++++++++++++++++-
 packages/engine/src/cron-runner.ts                 | 21 +++++-
 packages/engine/src/routine-runner.ts              | 11 +++-
 9 files changed, 255 insertions(+), 11 deletions(-)

Fusion-Task-Id: FN-7903

Fusion-Task-Lineage: c7eb4660-975d-4c5d-820e-0f1a3ac8b6a6

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 20:09:57 -07:00
gsxdsm
c110b72df1 FN-7902: persist thinkingLevel across mission/planning interview sessions 
Threads an optional per-session reasoning-effort (thinkingLevel) through mission planning and planning-mode agent sessions so the selected level survives draft reopen, session start, and agent rebuild.

- Add ThinkingLevel to DraftInputPayload/session state in planning.ts; persist and restore it in inputPayload alongside model overrides, threading it into createDraftSession, startExistingSession, createSessionWithAgent, initializeAgent, createPlanningAgent, and ensureSessionAgent as defaultThinkingLevel.
- Preserve thinkingLevel across draft syncs in ai-session-store.ts's updateDraft so it isn't erased when the model pair is unchanged.
- Extend mission-interview.ts session state/persistence and createMissionInterviewAgent/createMissionInterviewSession to accept and persist a validated thinkingLevel, defaulting the agent's reasoning effort from it.
- Validate thinkingLevel against THINKING_LEVELS in mission-routes.ts's POST /api/missions/interview/start and thread it through to the interview session.
- Update PlanningModeModal.tsx and MissionInterviewModal.tsx to surface and submit the selected thinking level; update legacy.ts API client to pass it through.
- Extend register-planning-subtask-routes.ts to accept/validate/forward thinkingLevel for subtask planning routes.
- Add regression coverage in mission-interview.test.ts, routes-planning.test.ts, and session-persistence-roundtrip.test.ts.
- Add changeset for @runfusion/fusion (minor): persisted thinking-level controls for Mission Interview and Planning mode.

Files changed:
 .changeset/quiet-dragons-think.md                  |   7 ++
 packages/dashboard/app/api/legacy.ts               |  14 ++-
 .../app/components/MissionInterviewModal.tsx       |  21 +++-
 .../dashboard/app/components/PlanningModeModal.tsx |  34 ++++--
 .../src/__tests__/mission-interview.test.ts        |  34 ++++++
 .../src/__tests__/routes-planning.test.ts          |  17 ++-
 .../session-persistence-roundtrip.test.ts          |  16 +++
 packages/dashboard/src/ai-session-store.ts         |  13 +-
 packages/dashboard/src/mission-interview.ts        |  30 ++++-
 packages/dashboard/src/mission-routes.ts           |  14 ++-
 packages/dashboard/src/planning.ts                 |  70 ++++++++---
 packages/dashboard/src/routes.ts                   |  11 +-
 .../src/routes/register-planning-subtask-routes.ts | 135 +++++++++++++++------
 13 files changed, 330 insertions(+), 86 deletions(-)

Fusion-Task-Id: FN-7902

Fusion-Task-Lineage: 751fb238-fd67-4870-975d-3b10dadde1a0

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 20:05:08 -07:00
gsxdsm
23cb061af2 FN-7898: add in-chat thinking-level control next to the attach button 
Adds a Brain-icon popover control in the chat composer that lets users change an existing session's reasoning-effort level mid-conversation, extending the existing create-time-only thinking-level picker.

- Add ChatThinkingLevelControl component (Brain-icon trigger + popover) wired into ChatView's direct-session composer, gated to non-CLI model-loop sessions only
- Extend PATCH /api/chat/sessions/:id to accept an optional thinkingLevel field, validated via existing validateThinkingLevel helper; null/empty string explicitly clears back to the inherited default, omitting the key leaves it untouched
- Add useChat().setSessionThinkingLevel hook method to call the new PATCH capability
- Add i18n strings (thinkingLevelButton) across all locales and regenerate packages/i18n/src/resources.d.ts
- Add changeset for @runfusion/fusion (minor) and update docs/dashboard-guide.md

Files changed:
 .changeset/fn-7898-chat-thinking-level-control.md  |   7 +
 docs/dashboard-guide.md                            |   2 +
 packages/dashboard/app/api/legacy.ts               |   4 +-
 .../app/components/ChatThinkingLevelControl.tsx    | 133 +++++++++++
 packages/dashboard/app/components/ChatView.css     |  73 ++++++
 packages/dashboard/app/components/ChatView.tsx     |  20 ++
 .../__tests__/ChatThinkingLevelControl.test.tsx    | 103 ++++++++
 .../__tests__/ChatView.message-edit.test.tsx       |   1 +
 .../components/__tests__/ChatView.test-harness.tsx |   1 +
 .../__tests__/ChatView.thinking-level.test.tsx     | 262 +++++++++++++++++++++
 .../dashboard/app/hooks/__tests__/useChat.test.ts  | 108 +++++++++
 packages/dashboard/app/hooks/useChat.ts            |  59 +++++
 .../dashboard/src/__tests__/chat-routes.test.ts    |  83 +++++++
 .../dashboard/src/routes/register-chat-routes.ts   |  35 ++-
 packages/i18n/locales/en/app.json                  |   1 +
 packages/i18n/locales/es/app.json                  |   1 +
 packages/i18n/locales/fr/app.json                  |   1 +
 packages/i18n/locales/ko/app.json                  |   1 +
 packages/i18n/locales/zh-CN/app.json               |   1 +
 packages/i18n/locales/zh-TW/app.json               |   1 +
 packages/i18n/src/resources.d.ts                   |  96 ++++++--
 21 files changed, 970 insertions(+), 23 deletions(-)

Fusion-Task-Id: FN-7898

Fusion-Task-Lineage: a052a6ef-d2d3-45af-9b92-221996780b1b

Co-authored-by: Fusion (runfusion.ai) <noreply@runfusion.ai>
 2026-07-12 20:00:50 -07:00