Rescues the QuickEntryBox suite by restoring jsdom globals instead of keeping it quarantined.
- Snapshot and restore QuickEntryBox viewport, visibility, matchMedia, and object URL descriptors after each test.
- Add a guard test that proves mutated jsdom globals return to their original descriptors.
- Remove the QuickEntryBox quarantine entries while documenting the rescue pattern.
Files changed:
docs/testing.md | 2 +
.../components/__tests__/QuickEntryBox.test.tsx | 50 ++++++++++++++++++++++
packages/dashboard/vitest.config.ts | 7 ++-
scripts/lib/test-quarantine.json | 5 ---
4 files changed, 55 insertions(+), 9 deletions(-)
Fusion-Task-Id: FN-6514
Fusion-Task-Lineage: 12f20ad8-19ee-4168-91a9-33193b1ab5f7
Expose declared fallback model lanes in project Settings while preserving workflow-scoped persistence.
- Load the default workflow definition before rendering Project Models lane controls.\n- Show fallback model dropdowns only when the workflow declares matching provider/model settings.\n- Persist and reset fallback lane edits through the Settings modal primary Save action.\n- Update settings documentation for default-workflow and global fallback model placement.
Files changed:\n docs/settings-reference.md | 40 +++---\n .../components/__tests__/SettingsModal.test.tsx | 135 ++++++++++++++++++++-\n .../settings/sections/ProjectModelsSection.tsx | 52 +++++++-\n 3 files changed, 204 insertions(+), 23 deletions(-)
Fusion-Task-Id: FN-6508
Fusion-Task-Lineage: 18dba2f9-658a-4b82-850f-3bbeb70341d3
Keep the Task Detail chat send glyph proportional to its touch target on desktop and mobile.
- Set the task chat send button icon sizing token to the larger spacing value.
- Preserve the existing desktop and mobile touch-target dimensions.
- Add CSS regression coverage for the send glyph and mobile sizing rules.
Files changed:
packages/dashboard/app/components/TaskChatTab.css | 6 ++++++
.../app/components/__tests__/TaskChatTab.test.tsx | 17 +++++++++++++++++
2 files changed, 23 insertions(+)
Fusion-Task-Id: FN-6507
Fusion-Task-Lineage: 619aef93-d225-443c-9cf2-41a59d794148
Keeps tablet task detail modals wide enough and within the viewport.\n\n- Reconcile the tablet overlay offset with modal max-height so actions remain visible.\n- Widen the tablet task detail modal to use more horizontal viewport space.\n- Add CSS-focused coverage for tablet sizing plus unchanged desktop and mobile guards.\n- Add a patch changeset for the published Fusion package.\n\nFiles changed:\n .changeset/fn-6500-tablet-task-detail-modal.md | 5 ++\n .../task-detail-modal-tablet-width.test.ts | 14 +++--\n .../dashboard/app/components/TaskDetailModal.css | 15 +++--\n ...etailModal.responsive-and-dependencies.test.tsx | 64 ++++++++++++++++++++++\n .../__tests__/core-modals-mobile.test.tsx | 5 +-\n 5 files changed, 94 insertions(+), 9 deletions(-)
Fusion-Task-Id: FN-6500
Fusion-Task-Lineage: 016a2839-29f9-48de-b644-a5fcd27b35e3
Keep tablet chat navigation available while software keyboards reduce the viewport.
- Stop auto-hiding the tablet chat sidebar when the keyboard opens.
- Bound the sidebar to its minimum width during tablet keyboard-open layout without overwriting the saved width.
- Cover visible and user-collapsed tablet sidebar behavior in mobile render tests.
- Add a patch changeset for the published Fusion package.
Files changed:
.changeset/fn-6494-tablet-chat-sidebar.md | 5 ++
packages/dashboard/app/components/ChatView.tsx | 34 +++--------
.../__tests__/ChatView.mobile-render.test.tsx | 70 ++++++++++++++++++++--
3 files changed, 79 insertions(+), 30 deletions(-)
Fusion-Task-Id: FN-6494
Fusion-Task-Lineage: 618cccc3-5102-4e9c-ac0e-6c7a44742940
The POST /chat/sessions/:id/messages writer (and cancel + isGenerating
enrichment) resolved its per-project ChatManager/ChatStore via
getOrCreateProjectStore, while the GET reader resolves via the engine-aware
resolveProjectChatContext. When those resolved to different store instances,
a sent message persisted to one store but the reload read from another, so
regular chat (ChatView) messages vanished after leaving and returning.
Quick Chat masked it by keeping its thread warm in memory (no server reload).
resolveScopedChatManager now resolves through resolveProjectChatContext so
writer and reader always share one store. Updated multi-project routing tests
to the corrected invariant and added an engine-aware regression test.
Relying on env(safe-area-inset-*) in remotely-served dashboard CSS was
unreliable (depends on device Chromium version reporting insets + the
service-worker-cacheable CSS being current). Configure
@capacitor-community/safe-area with detectViewportFitCoverChanges:false +
initialViewportFitCover:false so it pads the whole webview natively on every
Android device, guaranteeing the app never renders under the status/nav bars
regardless of CSS or webview version.
Android WebView returns 0 for env(safe-area-inset-top) by default, so the
CSS-only header inset had no effect on edge-to-edge devices (API 35+).
- Add @capacitor-community/safe-area@^7.0.0 (Capacitor 7 compatible); it
patches the webview so env(safe-area-inset-*) report real values. Enabled
natively (no JS init), so it works in remote/live mode too.
- capacitor.config.ts: declare SafeArea plugin (initialViewportFitCover).
- mobile-run-android.sh: idempotently enable EdgeToEdge in MainActivity after
cap add (android/ is generated/gitignored), required by the plugin.
Keep the Agents view focused when starting the selected agent from the dashboard TUI.
- Treat `s` as the selected-agent start command inside the Agents interactive view.
- Preserve `m` as the universal Main/status shortcut and keep the `s` alias outside Agents.
- Add regression coverage for Agents, non-Agents, status, and empty Agents shortcut behavior.
- Add a patch changeset for the published CLI package.
Files changed:
.changeset/fn-6491-tui-agents-start-key.md | 5 +
.../commands/dashboard-tui/__tests__/app.test.tsx | 102 ++++++++++++++++++++-
packages/cli/src/commands/dashboard-tui/app.tsx | 9 +-
3 files changed, 112 insertions(+), 4 deletions(-)
Fusion-Task-Id: FN-6491
Fusion-Task-Lineage: 82a9acab-6e4a-400c-80ee-54d051245b57
- Add scripts/mobile-run-android.sh + pnpm mobile:run:android (auto-detects
Android SDK + JDK 21, writes local.properties, reconnects network ADB
before deploy, supports remote backend via FUSION_SERVER_URL).
- Header.css: reserve env(safe-area-inset-top) so top app chrome no longer
draws under the OS status bar on edge-to-edge native shells (Capacitor
Android API 35+, iOS notch, PWA standalone). No-op on web/desktop.
- .gitignore: ignore generated packages/mobile/{ios,android} (stale paths
pointed at packages/dashboard/).
Replace raw dashboard RGB alpha colors with color-mix token expressions and guard the global CSS surface.
- Converted global theme shadow, state, mission, event, and theme-data alpha colors from raw rgba() to color-mix() expressions.
- Added CSS fixture loading for theme-data.css and a regression test banning raw rgb/rgba outside var() fallbacks across global app CSS.
- Documented the stricter dashboard styling rule for global and theme token CSS.
Files changed:
docs/dashboard-guide.md | 2 +-
.../__tests__/global-theme-css-no-raw-rgba.test.ts | 68 +++
packages/dashboard/app/public/theme-data.css | 573 +++++++++++----------
packages/dashboard/app/styles.css | 49 +-
packages/dashboard/app/test/cssFixture.ts | 7 +
5 files changed, 389 insertions(+), 310 deletions(-)
Fusion-Task-Id: FN-6489
Fusion-Task-Lineage: 7c466971-d15c-4382-a24b-1b32eea71974
Rescue the same-day core and CLI flaky quarantines by fixing their fixture seams instead of appeasing timeouts.
- Make the core concurrent-write lock helper release synchronously inside its child process so package load cannot delay the transient lock release.
- Close real CLI TaskStore fixtures before removing temp roots and switch mock cleanup to non-hoisted unmocking.
- Remove both test files from package quarantine excludes and clear the quarantine ledger while documenting the rescue pattern.
Files changed:
docs/testing.md | 2 ++
packages/cli/src/__tests__/extension-task-tools.test.ts | 17 +++++++++++++----
packages/cli/vitest.config.ts | 4 +++-
.../core/src/__tests__/store-concurrent-writes.test.ts | 8 +++++++-
packages/core/vitest.config.ts | 4 +++-
scripts/lib/test-quarantine.json | 13 +------------
6 files changed, 29 insertions(+), 19 deletions(-)
Fusion-Task-Id: FN-6486
Fusion-Task-Lineage: c90358c4-0549-4c68-8d17-2e2a336433b5
Quarantine the load-sensitive CLI extension task tools suite instead of widening Vitest timeouts.
- Add the extension task tools test file to the CLI Vitest quarantine exclude list.\n- Record the FN-6483 quarantine evidence in the deletion-ratchet ledger while preserving the existing core quarantine entry.\n\nFiles changed:\n packages/cli/vitest.config.ts | 5 +++++\n scripts/lib/test-quarantine.json | 5 +++++\n 2 files changed, 10 insertions(+)
Fusion-Task-Id: FN-6483
Fusion-Task-Lineage: 4a3be3b4-b2b4-4e05-afd0-564aa0d28c68
Require an out-of-repository operator signal before the release script can mutate version, publish, push, or tag.
- Add a reusable release authorization gate that allows dry-runs and blocks real releases without FUSION_RELEASE_AUTHORIZED.
- Invoke the gate in scripts/release.mjs before the first release mutation while preserving dry-run behavior.
- Cover blocked, authorized, dry-run, whitespace, TTY, and call-order behavior with script tests.
Files changed:
.../__tests__/release-authorization-gate.test.mjs | 70 ++++++++++++++++++++++
scripts/lib/release-authorization-gate.mjs | 30 ++++++++++
scripts/release.mjs | 26 +++++++-
3 files changed, 123 insertions(+), 3 deletions(-)
Fusion-Task-Id: FN-6480
Fusion-Task-Lineage: 5347552c-e395-4852-b389-6bbdba0e044e
- P1 (Greptile): a tool-use break-early turn released the warm connection
(inUse=false) while conn.prompt() was still pending, letting the next turn
launch a concurrent prompt on the same ACP session (protocol corruption).
keepWarm now requires !sawToolCall, so a tool-use turn tears the connection
down like the non-reuse path; only a clean stop turn (prompt fully resolved
before finish) keeps it warm. + test.
- buildBridgeEnv: treat a whitespace-only auth var as absent (v.trim()), so a
blank higher-preference token can't shadow a real lower-preference one and we
never forward a useless blank token. + test.
- Auth-forwarding tests: clear ambient auth vars in beforeEach so a runner-env
token can't shadow the case under test (CodeRabbit).
- Doc: clarify the allow-list never carries API keys by default; the single
FUSION_CLAUDE_ACP_FORWARD_AUTH opt-in (default OFF) is the only exception.
348/348 pass, tsc clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Keep a warm bridge connection + ACP session across turns of one
conversation (gated by FUSION_CLAUDE_ACP_REUSE=1, default OFF), so
multi-turn lanes skip the cold bridge/claude spawn and session/new
round-trip and send only the latest-turn delta (buildResumePrompt).
A stable router indirection serves each turn's handlers.
Addresses the adversarial review of the reuse path:
- P0: a warm-child death routes failure to the CURRENT owner turn via
router.fail, so a reuse turn fails fast instead of hanging until the
30-min inactivity timeout.
- P1: eviction is cache-identity-aware (evictCachedAcpConn only deletes
the map key when it still points at the entry), so a concurrent cold
turn / stale close handler / idle timer can't evict or kill a newer
live entry's child.
- P1: an empty resume delta cold-starts instead of issuing an empty
prompt that could hang.
- P2: a per-turn token drops cross-turn stray updates on the shared
warm connection.
- The idle reaper is unref'd so it never pins the process.
Default OFF → the cold path is functionally unchanged (reviewer-verified).
Adds multi-turn tests: reuse skips spawn+session/new, flag-off spawns
fresh, fail-fast on warm-child death, empty-resume cold fallback.
346/346 pass, tsc clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- P2: event-bridge handleMessageDelta now consumes cache_read/cache_creation
tokens (parity with handleMessageStart) — the OQ3 usage path carried them but
they were silently dropped, understating cost for cached turns.
- P2: validate the untrusted bridge usage payload — coerce each field to a
finite, non-negative number before forwarding, so a malformed value
(string/NaN/negative) can't corrupt totalTokens/cost.
- Tests: usage now asserts cache tokens + totalTokens; new cases for malformed
usage, tool-use turns reporting zero usage, the ANTHROPIC_AUTH_TOKEN middle
precedence, and that the auth token is read from process.env (never a
caller-supplied value — no token substitution).
- Doc: state the auth-forwarding exposure trade-off in the code comment.
acp-driver 13/13; event-bridge tests green; typecheck clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Item 2 (OQ3): capture PromptResponse.usage from conn.prompt and feed it into
the bridge before finish(), so ACP-path turns report token usage/cost instead
of always zero. Zero-when-absent is safe; tool-use (break-early) turns
inherently report zero (the prompt result never resolves).
- Item 3 (R17): opt-in headless credential delivery. When
FUSION_CLAUDE_ACP_FORWARD_AUTH=1, buildBridgeEnv forwards a SINGLE Claude auth
token (CLAUDE_CODE_OAUTH_TOKEN > ANTHROPIC_AUTH_TOKEN > ANTHROPIC_API_KEY) from
the operator's launch env so a detached daemon (no login Keychain) can
authenticate. Default OFF — the secure no-secrets posture is unchanged.
acp-driver tests 9/9 (usage + the three auth-opt-in cases); typecheck clean.
Remaining: item 1 (connection reuse / resume latency).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
GET /providers/claude-cli/status: asserts acp.{enabled,bridgeAvailable,active,
authFailed,authReason} reflect the FUSION_CLAUDE_ACP env + the bridge auth-failure
signal file, and that acp is inactive/clean when no bridge path is published.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Compound learning: the claude-code-cli-acp bridge returned 'Not logged in'
despite a working claude -p, due to (1) a too-thin spawn env (needs XDG_*/USER/
SHELL beyond HOME/PATH) and (2) macOS login-Keychain session isolation for
detached/headless processes. Six headless tasks misdiagnosed it as an upstream
gap. Cross-linked from the ACP runtime integration pattern doc.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- CodeRabbit: spinner class `spin` -> `animate-spin` (matches the card's other
Loader2 usages).
- CodeRabbit (major): tighten auth-failure detection so it only fires when the
WHOLE turn is the short "Not logged in" message (<=80 chars), not when a long
legitimate answer merely mentions the phrase — avoids false positives.
- CodeRabbit (major): expand the auth-signal test to assert the full invariant —
set on a not-logged-in turn, clear (unlink) on a real response, and NOT flag a
long answer that mentions the phrase.
pi-claude-cli acp-driver 5/5; typecheck clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Greptile P2: `acp.active` now reflects the ACTUAL dispatch determinant
(FUSION_CLAUDE_ACP, which includes the operator force-override), not the
experimental flag alone — so the status isn't misleading when forced on/off.
- CodeRabbit/Greptile P2: add FNXC:ClaudeAcp comments to the new code blocks
per the AGENTS.md greppable-comment convention.
Already fixed in the prior commit (daa37d08c): the P1 "sticky env" / latch
(applyClaudeAcpEnable now recomputes each call + FUSION_CLAUDE_ACP_FORCE
override) and the enable->disable-on-same-env regression test.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
When the bridged `claude` can't authenticate (detached daemon / no keychain),
the turn returns "Not logged in" instead of a real answer. Rather than silently
relay that, detect it and let the user choose.
- Driver: detect a "Not logged in"-only turn and write a cross-process signal
(fusion-acp-bridge-auth.json); a real response clears it (acp-driver test).
- Dashboard status: GET /providers/claude-cli/status reports
acp.authFailed + authReason from the signal.
- UI: the Claude CLI provider card shows an auth-failure banner with
"Use claude -p" (sets experimentalFeatures.claudeCliAcp=false) and
"I fixed auth — re-test", plus a fix hint (run `claude` to log in).
- Enable resolution now recomputes each call with an operator force-override
(FUSION_CLAUDE_ACP_FORCE), so the "Use -p" fallback takes effect on the next
turn — no restart. claude-acp-enable tests updated.
pi-claude-cli + engine tests green; dashboard typecheck clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
GET /providers/claude-cli/status now returns an `acp` block:
{ enabled (experimental flag), bridgeAvailable (KTD10 published path), active
(enabled && acpEnabled && bridgeAvailable) } so operators can see whether Claude
CLI is routing through the ACP bridge vs `claude -p` — important for the
default-on rollout. Additive; typechecks clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace the manual FUSION_CLAUDE_ACP env enable with an experimental feature
switch. `experimentalFeatures.claudeCliAcp` is ON by default (off only when
explicitly set false); the engine translates it into the FUSION_CLAUDE_ACP
dispatch the pi-claude-cli provider reads, at registerExtensionProviders time.
- Still fail-closed: with no bridge path published (acp-runtime plugin absent),
the provider falls back to `claude -p`.
- Explicit FUSION_CLAUDE_ACP env always wins (operator / test override).
- New testable helper claude-acp-enable.ts (6/6 tests); flag documented in the
core experimentalFeatures doc.
So with the acp-runtime plugin installed, Claude CLI now routes through the ACP
bridge by default; set experimentalFeatures.claudeCliAcp=false to force `-p`.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Two-reviewer pass (security + architecture) on KTD10 + the full Route A
increment: no code defects, no P0, merge-ready as a dormant increment. Applying
the P1 follow-ups:
- Add the feature changeset (@runfusion/fusion minor) — the one convention gap.
- KTD10 tests: fail-closed (bridge not resolved -> env stays unset -> -p) and
idempotency (second onLoad keeps the first published path).
- Document the two intentional, parallel MCP-forwarding paths (U10 engine-adapter
vs U11 provider-driver) so nobody double-forwards, and the known
ACP-path-token-usage=0 residual so U12 doesn't treat it as a bug.
Reviewers confirmed: dormancy invariant holds end-to-end (nothing sets
FUSION_CLAUDE_ACP=1; both flag+path required; -p is the default); OAuth pi path
untouched. 206/206 plugin tests, 333/333 pi-claude-cli tests, typecheck clean.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
