- Add onboarding project-mapping API contracts and dashboard UI wiring in AddNodeModal/NodesView
- Persist node-to-project mappings with rollback-safe failure handling and expanded hook/API test coverage
- Document node onboarding mapping behavior in architecture and multi-project docs
- Add tsup bundle asset gates for openclaw bridge and droid runtime, with changesets and bundle output tests
Fusion-Task-Id: FN-3505
Stages the droid runtime as a bundled plugin in the CLI's tsup configuration, gates the bundle bootstrap on droid asset availability, and adds tests asserting correct staging behavior in bundle output.
Fusion-Task-Id: FN-3718
Stabilizes the OpenClaw bundle by hardening the asset copy logic and gating the bootstrap flow on the openclaw bridge asset being present, with regression tests covering the new bundle-output helpers and a patch changeset for `@runfusion/fusion`.
Fusion-Task-Id: FN-3717
This merge introduces three major features: a full path-mapping API layer across core, dashboard, and desktop (FN-3504, four commits), shell multi-profile support with a new NativeShellConnectionManager component and connection manager onboarding flow (FN-3403), and migration of the WhatsApp plugin
Fusion-Task-Id: FN-3504
Merges shell multi-profile support for desktop (FN-3403) — spanning new connection-manager flows, shell onboarding interoperability, mobile connection profiles, desktop IPC/shell-settings plumbing, and tokenized titlebar styles — with a secondary migration of the WhatsApp plugin to the Baileys pairi
Fusion-Task-Id: FN-3403
Cleans up legacy webhook key references from the schema test file, removing outdated assertions that referenced deprecated fields.
Fusion-Task-Id: FN-3713
The merge lands three major features: a permanent-agent gating system (FN-3560, 6 steps) that classifies and enforces tool access policies for permanent agents in the PI extension, with full test coverage and updated agent docs; an OpenClaw MCP bridge (FN-3717) adding MCP config, schema server, and
Fusion-Task-Id: FN-3560
Adds a full evals dashboard view (FN-3394) with a client hook, server-side read routes, and navigation integration across the Header and mobile nav, backed by comprehensive tests. Wires Hermes chat message tools with mailbox behavior documentation (FN-3620), and extends the agent heartbeat and tools
Fusion-Task-Id: FN-3394
Merges FN-3620, completing the Hermes chat mailbox integration with wired message tools, locked sender/recipient contract tests, and mailbox behavior documentation, plus a small CSS polish adding `focus-visible` styles and token radius to the render toggle. A changeset for FN-3710 (cluster task ID o
Fusion-Task-Id: FN-3620
Adds an agent workspace memory reader so agents can access workspace context, and extends the planning mode modal with priority controls for task routes and subtasks (FN-3716, FN-3718). Minor CSS token and mobile focus fixes are also included (FN-3710).
Fusion-Task-Id: FN-3710
Merged branch lands two features: FN-3718 wires workspace memory into the agent instruction pipeline — adding a reader helper, injecting workspace memory into agent instructions and identity snapshots, and documenting the resolution order — and FN-3716 adds planning mode priority controls for task r
Fusion-Task-Id: FN-3718
Adds per-subtask priority controls to planning mode with three implementation steps (route priority support, UI controls, and design-token polish), gates research tools behind an experimental feature flag with a new core infrastructure module, and introduces a WhatsApp chat plugin with discovery and
Fusion-Task-Id: FN-3716
The merge introduces a research tool surface gating mechanism: shared availability helpers in core and engine, applied to the executor and triage agent so research prompts and tool exposure are gated behind experimental-feature flags, with tests covering the new logic.
Fusion-Task-Id: FN-3706
Adds a WhatsApp chat plugin (`plugins/fusion-plugin-whatsapp-chat/`) as the first bundled plugin, including the plugin package with UI rendering, dashboard plugin manager integration, CLI bundling support, and plugin route runtime context. Also updates documentation to cover plugin management and ap
Fusion-Task-Id: FN-3714
- Always apply configured-provider filter (even when empty set)
- Convert getConfiguredProviderNames to async (fs/promises)
- Validate model overrides as non-empty provider+model pair
- Clear stale persisted insight model when no longer available
- Replace hardcoded max-height with design token in overflow menu
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Wire up CustomModelDropdown in InsightsView with favorites support:
favoriteProviders, favoriteModels, and toggle callbacks
- Auto-promote resolved planning provider as implicit favorite when
no explicit favorites are set
- Filter /models API to only return providers with credentials in
Fusion's own auth stores (primary + legacy .pi + models.json),
excluding supplemental stores (Codex CLI, Claude Code, env vars)
that surface unconfigured providers
- Still allow enabled extension providers (Claude CLI, Droid CLI,
llama.cpp) when their settings toggles are on
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
InsightsView accepted a models prop that was never passed from App.tsx.
Now fetches models via fetchModels() on mount (same pattern as ChatView),
falling back to the prop if provided.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
getSettingsFast() and getSettings() spread project-level settings over
global settings without filtering by scope. A stale project-level
experimentalFeatures: {} in SQLite would override the global
experimentalFeatures: { insights: true, ... }, making all experimental
features invisible in the dashboard.
Now both methods strip global-only keys from project settings before
the merge, matching the behavior of getSettingsByScopeFast().
Also adds max-height and overflow-y to the overflow dropdown menu.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Stores modelProvider/modelId in the run's inputMetadata.metadata at
creation time, then recovers it in the retry handler so the user's
explicit model choice survives a transient failure and retry.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Guard against malformed localStorage values (e.g. "provider/") by
validating both parts are non-empty before sending to the server
- Only wire settings fallback when both provider AND modelId are
explicitly provided, preventing mismatched provider/modelId combos
Addresses ISSUE-002 and ISSUE-003 from code quality review.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adds a Settings gear button to the InsightsView action bar that toggles a
collapsible model configuration row. The row contains a CustomModelDropdown
allowing users to override the insight generation model. Selections persist
to localStorage under the "fusion-insight-model" key. A yellow indicator dot
appears on the gear when a non-default model is active.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adds documentation for the approval architecture and runtime policy contract to `docs/architecture.md` and `docs/agents.md` (FN-3712).
Fusion-Task-Id: FN-3712
Merged seven commits that introduce a runtime action gate for task execution (FN-3547) — adding action classification, git-based heuristics for branch/remote gating, and per-step session enforcement — with tests covering the gate logic, heartbeat integration, and PI agent creation. Also landed FN-37
Fusion-Task-Id: FN-3547
Consolidated graph position storage into a canonical shared helper (`projectStorage.ts`) with hardened persistence logic and removed the duplicate `scopedStorage` module. Updated integration tests and documented the canonical storage approach. A smaller fix (`FN-3628`) addressed mobile touch targets
Fusion-Task-Id: FN-3711
The merge brings a mobile accessibility fix to ChatView, improving touch target sizing and aligning CSS token usage with the dashboard's design system conventions.
Fusion-Task-Id: FN-3628
Merges a substantial cross-cutting batch: approval request persistence with a SQLite store (FN-3546), branch/baseBranch fields added to the task schema (FN-3429), project-node path mapping APIs and storage layer (FN-3428/FN-3503), draft-to-form unification and interview close/confirmation boundaries
Fusion-Task-Id: FN-3557
Merged branches introduce an approval-request persistence layer (FN-3546) with a new `ApprovalRequestStore` in core, schema v68 with audit history support, and updated dashboard task/API tests, alongside a minor task-schema extension (FN-3429) adding `branch`/`baseBranch` fields to the task model an
Fusion-Task-Id: FN-3546
Adds a changeset for feature FN-3429 to track the branch/baseBranch task fields, marking this work for the next `@runfusion/fusion` release.
Fusion-Task-Id: FN-3429
Merges FN-3428's branch-field contract and merge target override/default behavior tests across core, dashboard, and engine, plus a companion fix that sweeps subsumed autostash orphans and surfaces outcomes on the task feed. Adds project-node path mapping persistence APIs and schema (FN-3503), new ag
Fusion-Task-Id: FN-3428
Autostash entries from prior runs piled up indefinitely (50+ on a single
working tree) because every silent restore failure (apply hard-fails on
untracked-overwrite, transient git error, etc.) leaves a permanent stash
and the warn-only behavior trains developers to ignore the warnings.
- Add `sweepAutostashOrphans`, called at merge entry: classifies each
orphan by diffing its stashed paths against HEAD. If every path is
byte-identical to HEAD the dev's work has already landed, so drop the
stash automatically. Live orphans (real lost work) get loud warnings
on both the engine log and the task feed.
- Surface every restore terminal outcome (`restored`, `failed`,
`conflict-needs-manual` from each path) via `store.logEntry` so devs
can see in `fn task show` what happened to their stash without
grepping engine logs.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds project-node path mapping infrastructure in the core database layer (schema, migration backfill, and persistence APIs in `central-core.ts` and `central-db.ts`) with accompanying documentation. Also introduces a review-before-apply draft gate for agent onboarding and the agent detail view's mail
Fusion-Task-Id: FN-3503
The leak guard in scripts/check-test-isolation.mjs reported
fusion-test-home-root-* dirs created by scripts/test-changed.mjs as
leaks when transient EBUSY on /var/folders prevented cleanup or the
baseline file got rotated. Track every HOME basename the script mints
and pass it via FUSION_TEST_ISOLATION_IGNORE_NAMES so the check
allow-lists them unconditionally. Surface cleanup rm failures via
console.warn instead of swallowing.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The merge adds a mail tab to the agent detail view in the dashboard (with wiring and documentation), implements a review-before-apply draft gate in the experimental agent onboarding modal, and introduces an opt-in web builtin tool for AI agents with engine and CLI reference updates.
Fusion-Task-Id: FN-3701
Adds a review-before-apply flow to the Experimental Agent Onboarding Modal with an explicit draft review gate, updating the modal component and its co-located CSS, plus test coverage. Also corrects the focus ring token used in the ChatView session delete action.
Fusion-Task-Id: FN-3465
This merge adds an agent permission policy model with persistence (FN-3545), an explicit draft review gate for task workflow steps (FN-3556), web builtin tool opt-in for AI agents (FN-3713), an inline fast mode toggle wired into peer execution (FN-3711), and a retry mechanism for overlapping task ID
Fusion-Task-Id: FN-3556
Added agent permission policy model with persistence in core, and wired web-builtin tool opt-in enabling planning and synthesis web tools in the dashboard with corresponding docs and a changeset. Also added an inline fast-mode toggle wired into peer executor state, retry logic for cluster task-ID ov
Fusion-Task-Id: FN-3713
Introduces a multi-agent permission policy model in `@fusion/core` (types, store persistence, API exposure) with corresponding dashboard UI wiring in the agent onboarding modal and task detail modal, plus a distributed task ID overlap retry mechanism for cluster creation, an inline fast mode toggle
Fusion-Task-Id: FN-3545
This merge introduces several dashboard and engine improvements: agent inbox wake controls (FN-3087, FN-3588) with immediate message triggers; separated plugin lifecycle from setup probe state (FN-3704); assigned-agent triage inheritance in the engine (FN-3703); inline fast mode toggle in the task d
Fusion-Task-Id: FN-3711
Adds retry logic for distributed task ID overlap conflicts across the cluster, hardening the ID allocator with exponential-backoff recovery, wiring the retry into task workflow registration, and documenting the behavior in architecture docs. Minor CSS tokenization on the Mailbox and Plugin Manager m
Fusion-Task-Id: FN-3710
The merge delivers four distinct features: a copy-to-clipboard action on provider responses in the chat bubble (FN-3712, with styling and tests across ChatView), richer agent onboarding drafts documented and wired into the dashboard onboarding modal (FN-3555), assigned-agent triage inheritance so ag
Fusion-Task-Id: FN-3712
Documents richer onboarding drafts in agents.md and updates the dashboard README with a small addition, completing Step 4 of the onboarding feature.
Fusion-Task-Id: FN-3555
Implements assigned-agent triage inheritance (FN-3703), allowing triage logic to be delegated to the assigned agent rather than always routing through the original owner, with test coverage and documentation. Also adds immediate wake controls for agent inbox and message API (FN-3087), wires shared s
Fusion-Task-Id: FN-3703
Merges peer exchange shared state wiring (FN-3464) into the engine with updated tests, adds immediate wake controls for agent inbox and message API (FN-3087), and restores the narrow logs/system mouse auto-toggle policy with CLI documentation (FN-3708). Dependency graph plugin receives test and high
Fusion-Task-Id: FN-3464
