v0.2.0 - New Assets Modules + some bug fixes

[TruthSeeker04]

LogCore OS v0.2.0 — The Assets Release

This release introduces the Assets module — track anything you own in a private, template-driven object tree — plus the Automation Inbox for reviewing workflow results, a safer default AI chat mode, and a long list of stability fixes.

---

🗂️ New module: Assets

Track anything ownable — land parcels, vehicles, equipment, tools — as a nestable tree (subdivision → parcels → …), available in both personal and business workspaces.

Templates

• Define each object type's premade fields (text / number / date / boolean / select, with optional defaults), icons via a built-in emoji picker
• Anyone can create their own templates and share them with specific people or whole roles; admins manage global templates and can restrict them by role
• A default 📁 Folder template ships out of the box — just a name and notes, for grouping assets without building a template first

A clean, read-first view

• Clicking an asset opens a readable overview — fields at a glance, attachments, the items inside it, and linked tasks — with an ✎ Edit button to switch into the editor (shown only if you can edit)

Sharing that asks first

• Sharing an asset or template sends each person an Accept/Decline request — it only appears for them once they accept, and they can leave a share later
• Share to Team, Household, a role, or a specific person; defaults to "everything inside" with a "this one only" option; new assets created inside a shared one automatically join the same audience
• Hide specific assets from selected users or whole roles (e.g. role:crew) — role hides follow the role, including people assigned to it later

Contribute access — built for employees and crew

• A middle access level where you pick exactly what a person can do: which fields they may change (e.g. just Status) and what they may add (comments, photos/PDFs, items inside)
• Contributors on Team/Household assets get the same picked-capability access without full team management rights
• Per-person grants always beat group grants, so restricting one member of a shared group actually works
• Contribute users work from the clean asset view (quick status dropdown, only their granted fields, comment box) — they never see the full editor

Comments

• An attributed, append-only note log per asset ("gate fixed, invoice sent") that can't overwrite anyone else's text
• Everyone with edit access is notified on new comments — the bell notification has a View → button (and pushes deep-link) straight to that asset
• Audit-style by design: only an admin can delete a comment; edit-level users can turn comments off for everyone on an asset (data kept); anyone can collapse the section for themselves
• 🔔 Per-asset notification mute covering that asset and everything inside it — a personal preference that affects no one else

And the rest

• Admins can convert an asset tree into a shared Team/Household object that survives user account deletion
• Archive a single asset or its whole subtree (you're asked which); per-asset change history; photo/PDF attachments
• Link tasks to assets from either side (task form asset picker; "＋ Task for this asset" in the asset editor)
• Search bar and filters (owned / shared / pool / by type) over the real foldered tree; move assets with a tree picker
• AI chat can list, search, create, update, move, and archive assets (writes require your approval); admins can manage templates by chat
• n8n automation API: token-authenticated endpoints to list, create, update, and comment on assets from workflows; token managed in Admin → n8n

📥 Automation Inbox

• Workflows (n8n) can post structured, reviewable results into LogCore — land leads, price alerts, anything — reviewed from a new Inbox view on the Automations page with one-tap actions: Interested / Pass / Offer Made / Closed (every action records who took it)
• Named inboxes: create as many as you need, route each workflow's output to the right one, and pick per inbox who gets notified and who may act (admins always can); unrouted results land in an auto-created General inbox
• New items send one batched notification (bell + push) with a View → button that jumps straight to the right inbox — switching workspace if needed
• Duplicate protection: re-running a workflow never re-adds items it already submitted, and workflows can ask LogCore what it has already seen before spending AI credits qualifying old listings
• Works in both workspaces: business inboxes are shared with the team; personal inboxes are yours

✅ Approve-edits chat mode (new default)

• The AI now pauses before any data change and shows an approval card; reads run freely
• Plan / Auto / Research modes unchanged and still selectable

Changed

• Goal-type tasks now live on the Goals page only — removed from the Tasks page, the dashboard's Top 3 and Due Today, and the morning digest (they still show on the calendar when dated; goal-drift check-ins are unchanged)

Fixed

• Per-person asset permissions now beat group permissions — tightening one member's access takes effect even when a household/team-wide share exists
• A per-person contributor entry now also restricts users holding the "Can manage" pool grant (admins are never restricted)
• Saving a just-created Team/Household asset no longer fails with a "Pool assets are workspace-visible" error
• No longer randomly logged out — a transient/background 401 used to clear the session and bounce you to login; the app now re-verifies first
• Sign-out fully resets the theme — the login page always shows brand orange and no longer keeps the last user's background or accent
• Login banner now fades in smoothly and is preloaded/cached so it appears instantly
• Fixed a crash ("Something went wrong") when the notifications list came back in an unexpected shape; the asset editor no longer crashes on unusual history data; the error screen gained a Reload button
• Mobile: asset and template editors no longer extend under the phone status bar (safe-area-aware modals)
• Saved chat archives no longer lose multi-line AI responses, and long AI responses (over 5,000 chars) now auto-save correctly
• Proactive notifications injected into chat no longer break sending messages or create junk chat archives
• The AI agent now resolves household member names on task assignment (first-name matching; asks when ambiguous) instead of writing raw strings

---

Updating: Admin → Updates, or run bash docker/update.sh on the host. Brain schema migrations run automatically at startup.

Full changelog: v0.1.0...v0.2.0

---

This discussion was created from the release v0.2.0 - New Assets Modules + some bug fixes.