598 results found

1. make it easy as possible to run Lumo from a home server - its a win for all but especially Proton

   Proton allows you to download and run your own version, but I just read the instructions and its not trivial. Its in the interest to Proton to do so. The user takes on costs for power, space, and cooling. An analogy: a german telcom created a virtualized TV set top box. The TV interface was streamed and effectively ran like a video game. The home only required a simple USB stick that ran DRM and some controls. This cost about 30 euro and the user would self install, just plug it into their smart TV. This replaced a 100 Euro set top box and a service call to do the install (150 Euro). Seems simple no? What they didnt consider is the home owner pays the power and space of the STB, this was a net new cost the Telco had to take on. And you have to build out for peak demand even when no one is watching. It was a massive failure. This is why distributed AI "on the edge" of the network makes financial sense. Let the user take on the power and space costs. The benefit to the user is complete sovereignty and privacy. But it will never be popular unless you make trivially easy. Do a deal with some hardware vendor to sell preconfigured rasberry pi boxes with a simple wizard interface to configure it. Provide the home version with security updates etc.

   Proton allows you to download and run your own version, but I just read the instructions and its not trivial. Its in the interest to Proton to do so. The user takes on costs for power, space, and cooling. An analogy: a german telcom created a virtualized TV set top box. The TV interface was streamed and effectively ran like a video game. The home only required a simple USB stick that ran DRM and some controls. This cost about 30 euro and the user would self install, just plug it into their smart TV. This replaced a 100 Euro… more

   4 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  Changes to existing features  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

2. Add an incognito mode to not include memories for unbiased chats

   Lumo is great and the memory feature is great. However sometimes it would be nice to have a chat without it having all that extra memory context to have a simple unbiased, impartially neutral chat for times where all the memory content is making things "too" personalised with the angles it takes.

   It would also come in handy when someone is next to you and asks "Can you ask Lumo about XYZ" and then you search. With an incognito feature, if you're asking it questions about someone else's issues it won't blurt out for everyone reading things like "It's not that different than your [insert personal PRIVATE topic here] issue." because it wouldn't know any of your prior history/memory for that new chat.

   Lumo is great and the memory feature is great. However sometimes it would be nice to have a chat without it having all that extra memory context to have a simple unbiased, impartially neutral chat for times where all the memory content is making things "too" personalised with the angles it takes.

   It would also come in handy when someone is next to you and asks "Can you ask Lumo about XYZ" and then you search. With an incognito feature, if you're asking it questions about someone else's issues it won't blurt out for everyone reading things like "It's not… more

   4 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  New feature  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

3. Dark app icon for Lumo

   All the other Proton apps have an icon for light and dark mode. I think it would be great to have a dark icon for Lumo since it only has a light icon.

   27 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   2 comments  ·  New feature  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

4. 🤖 Autonomous Project Builder with Local File System Access

   🤖 Autonomous Project Builder with Local File System Access

   Summary

   Transform Lumo from a code-writing assistant into an autonomous project builder that can read, write, organize, and execute code directly on user's local filesystem with proper security controls.

   Problem Statement

   Current AI coding assistants (including current Lumo) operate in a "chat-only" paradigm:
   - User: "Build a Flask API with authentication"
   - Lumo: Returns code blocks in chat
   - User: Manually copies files, creates directories, runs tests, configures environment
   - Result: Friction, human error, context drift, repetitive work

   Real engineering work requires:
   1. Creating directory structures
   2. Writing multiple files with correct paths
   3. Running commands/tests locally
   4. Debugging errors from actual execution
   5. Iterating based on real runtime behavior

   Without file system integration, Lumo cannot close the loop from "design" to "deployed working code."

   Proposed Solution

   Core Capability: Secure Local Workspace Access

   Mode 1: Read-Only Analysis (always enabled)
   - Analyze existing project structure
   - Identify dependencies, bugs, gaps
   - Suggest improvements with precise file locations

   Mode 2: Controlled Write Access (opt-in, sandboxed)

   User Permission Required For: ├── Creating new files/folders ├── Modifying existing files ├── Running shell commands (explicit allowlist) ├── Installing packages (npm/pip/cargo with review) └── Committing to Git (preview before commit)

   Mode 3: Autonomous Execution (advanced, explicit opt-in)
   - Lumo builds complete features end-to-end
   - Runs tests automatically
   - Fixes errors without waiting for approval (within defined scope)
   - Reports back: "Built X, ran Y tests, Z passed, here's the diff"

   Security Model (Privacy-First Design)

   Security_Guardrails:
     - Scope_Limit: Only access explicitly designated workspace folders
     - Command_Allowlist: Pre-approved commands (git, pytest, npm test, etc.)
     - Network_Isolation: No outbound calls without user approval
     - Audit_Log: Every action logged with diff preview before/after
     - Rollback: Atomic writes + version history (undo any change)
     - E2EE: All local data encrypted at rest
   User Interface Patterns
   
   Pattern A: Sidebar Integration
   [Lumo Chat] | [Project Explorer] | [Terminal Output]
   ────────────│─────────────────│──────────────────
   Code dialog │ Files Lumo touched│ Live command output
               │ (read-only view) │ (with error highlighting)
   
   Pattern B: Progress Tracking
   Building Adrenalina Prime Module:
   ✓ Created /src/core/metrics_collector.py
   ✓ Created /src/core/energy_engine.py
   ✓ Generated tests (87% coverage)
   ⏳ Running pytest...
   ✗ Fixing failure in test_crossbot_consensus.py
   ✓ All tests passing (92%)
   → Ready to deploy? [Yes/Review Changes]
   Example Workflow Comparison
   Traditional Chat    With Local File Access
   Paste code → copy to files    Lumo writes files directly
   Manual env setup    pip install -r requirements.txt (auto)
   Run tests manually  Tests auto-run after each change
   Find bug in console Lumo reads error, fixes code, re-tests
   Git commit manually Diff preview → one-click commit
   User Stories
   
       "As a solo developer, I want Lumo to scaffold my entire project structure so I don't waste hours on boilerplate."
   
       "As a senior engineer, I want Lumo to fix failing tests autonomously within my project folder so iteration speed is faster."
   
       "As a privacy-conscious user, I want to define exact folder boundaries Lumo can touch and see every change before it's applied."
   
       "As a team member, I want Lumo to review my git diff and explain why each file changed."
   
   Technical Requirements
   Backend Infrastructure
   
       Local agent process (runs on user machine, not cloud)
       File system watchers for real-time sync
       Sandbox container for untrusted operations
       SQLite/encrypted store for version history
   
   Integration Points
   Existing Lumo Features          New Requirements
   ────────────────────           ──────────────────
   Persistent Memory        →     Project-scoped workspace binding
   File Upload              →     Bidirectional file sync
   Custom Lumos             →     Role-based permission profiles
   Web Search               →     Package registry lookups (local)
   API Contract Example
   class LumoWorkspace:
       def list_files(self, path: str) -> List[str]:  # Read
       def read_file(self, path: str) -> str:         # Read
       def write_file(self, path: str, content: str) -> bool:  # Needs approval
       def run_command(self, cmd: str) -> CommandResult:       # Allowlist only
       def get_diff(self, path: str) -> str:                  # Preview
   Success Metrics
   Metric  Target
   Time-to-working-code (vs manual)    >70% reduction
   Session-to-deployment cycle time    <15 minutes for small features
   User trust (permission acceptance rate) >80% opt-in within 6 months
   Error resolution speed  3x faster than manual debugging
   Risks & Mitigations
   Risk    Severity    Mitigation
   Malicious prompt injection writing arbitrary files  🔴 Critical   Sandboxed execution; explicit approval for all writes
   Data exfiltration via file uploads  🔴 Critical   No outbound network calls without audit log + approval
   Accidental corruption of production code    🟡 High   Atomic writes + automatic backups + undo history
   Performance impact on large codebases   🟢 Low    Incremental indexing; lazy loading
   Competitive Landscape
   Platform    File Write  Auto-Test   Permission Model    Privacy
   Cursor  ✅ ⚠️ Partial  ⚠️ Basic    ❌ Cloud-first
   GitHub Copilot Workspace    ✅ ⚠️ Limited  ❌ None    ❌ Microsoft cloud
   Windsurf    ✅ ⚠️ Partial  ❌ No permissions  ⚠️ Hybrid
   Lumo (proposed) ✅ ✅ ✅ Granular    ✅ E2EE local
   
   Differentiation: No competitor offers privacy-preserving autonomous execution with granular user control. This would be Proton's unique value proposition.
   Priority
   
   Critical — This transforms Lumo from "chatbot" to "engineering platform"
   Effort
   
   High — Requires local agent infrastructure + security review + UX design
   Timeline Estimate
   
   6-8 sprints for MVP (read-only analysis → controlled writes → autonomous mode)
   Related Features
   
       Persistent Memory (needs project-scoped binding)
       Custom Lumos (can define permission profiles)
       File Analysis (upgrade to bidirectional sync)
       Proton Desktop App (could host local agent process)
   
   Engineer's Note
   
   I'm proposing this after building complex systems for 25+ years. The bottleneck isn't code generation anymore — it's closing the loop between idea and running system.
   
   Every time I have to:
   
       Copy code from chat to files
       Manually create directories
       Run commands I asked Lumo about
       Debug errors Lumo couldn't see
   
   That's 30-60 minutes wasted per feature. With secure local access, that becomes 5 minutes.
   
   Privacy isn't negotiable. Everything stays on-device. No cloud processing. End-to-end encryption. This is Proton's DNA.
   
   ---
   
   ## **Sekcje dodatkowe (opcjonalne):**
   
   **Dodatkowy komentarz przy zgłoszeniu:**
   
   This is not theoretical — I've prototyped local agent systems that can build and test code autonomously. The technology exists. What's missing is a privacy-preserving implementation that engineers trust.
   
   Key technical decisions already validated:
   
       Atomic writes prevent data corruption
       Sandboxed containers limit damage scope
       Version history enables rollback to any point
       Allowlist-based command execution prevents abuse
   
   The question isn't "can we build this?" It's "will Proton lead on privacy-preserving AI engineering tools?"
   
   **Tagi sugerowane:** `productivity`, `security`, `developer-tools`, `automation`
   

   🤖 Autonomous Project Builder with Local File System Access

   Summary

   Transform Lumo from a code-writing assistant into an autonomous project builder that can read, write, organize, and execute code directly on user's local filesystem with proper security controls.

   Problem Statement

   Current AI coding assistants (including current Lumo) operate in a "chat-only" paradigm:
   - User: "Build a Flask API with authentication"
   - Lumo: Returns code blocks in chat
   - User: Manually copies files, creates directories, runs tests, configures environment
   - Result: Friction, human error, context drift, repetitive work

   Real engineering work requires:
   1. Creating directory structures
   2. Writing multiple files with… more

   3 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  New feature  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

5. 🧠 Advanced Persistent Project Memory & Engineering Partner Mode

   🧠 Advanced Persistent Project Memory & Engineering Partner Mode

   Summary

   Enable Lumo to maintain long-term project context across sessions and operate as a proactive engineering partner rather than reactive assistant.

   Problem Statement

   Current conversation memory is session-bound. Each new chat requires re-establishing context about:
   - Project architecture decisions and trade-offs
   - Technology stack specifics (local models, frameworks, deployment targets)
   - Previous iterations and lessons learned
   - Ongoing development challenges

   This creates friction for complex, multi-session projects and forces users to constantly re-explain context.

   Proposed Solution

   Feature 1: Project-Specific Memory Stores

   Description: Allow users to create and manage named "project memory" objects that persist across separate conversations.

   Key Capabilities:
   - Store: Architecture diagrams, tech stack details, API contracts, codebase structure
   - Query: Natural language retrieval of project-specific knowledge
   - Update: Incremental additions as project evolves
   - Export/Import: Share project contexts between team members

   Technical Implementation:
   ```yaml
   ProjectMemory:
   - Name: "adrenalina-prime"
   - Created: timestamp
   - Tags: ["engineering", "agent-system", "local-llm"]
   - ContextBlocks:
   - Type: "architecture" | "api-contract" | "tech-stack" | "decision-log"
   - Content: markdown/text
   - Last_Modified: timestamp
   Feature 2: Engineering Partner Mode

   Description: Toggle switch for interaction style optimization.
   Standard Mode Engineering Partner Mode
   Reactive Q&A Proactive gap identification
   Generic responses Stack-aware suggestions (e.g., knows Ollama vs OpenAPI constraints)
   No state awareness Remembers previous session's blockers
   Neutral tone Direct, technical communication (zero sugar-coating)
   Feature 3: Contextual Code Awareness

   Description: When uploading code files, Lumo should build internal map of:

   File dependencies and imports
   Function signatures and interfaces
   Testing coverage and gaps
   Potential security vulnerabilities based on patterns
   

   Integration Points:

   Auto-detect language/framework from file extensions
   Cross-reference with known local model capabilities (e.g., deepseek-coder:6.7b context window limits)
   Suggest tests before implementation, not after
   

   User Stories

   "As a developer working on a multi-week project, I want to resume conversation without explaining what we built last week."
   
   "As an engineer, I want Lumo to notice architectural inconsistencies between my code and stated design goals."
   
   "As a team lead, I want to share project context with junior developers via exported memory bundles."
   

   Success Metrics
   Metric Target
   Reduction in context-setting time >60% decrease
   Multi-session project completion rate +40% increase
   User satisfaction (Engineering users) NPS +25 points
   Memory feature adoption >50% of Lumo Plus users within 6 months
   Risks & Mitigations
   Risk Mitigation
   Privacy concerns with persistent data All memory encrypted end-to-end; opt-in only
   Memory bloat over time Automatic summarization; user-controlled cleanup
   Inaccurate recall Version history with diff view; confidence scoring on retrieved context
   Competitive Differentiation

   Other AI assistants (Claude, ChatGPT, etc.) offer:

   ✅ Long context windows
   ❌ Project-scoped memory persistence
   ❌ Role-based interaction modes
   ❌ Architecture-aware code review
   

   Lumo's opportunity: Become the first privacy-focused AI with persistent project memory + professional mode optimization.

   Priority: High Effort: Medium (requires backend memory infrastructure + UI toggles) Timeline Estimate: 3-4 sprints for MVP

   Related Features:

   Custom Lumos (already exists, but lacks memory integration)
   Persistent Memory (exists, but not project-scoped)
   File Analysis (exists, but lacks cross-session code mapping)
   

   🧠 Advanced Persistent Project Memory & Engineering Partner Mode

   Summary

   Enable Lumo to maintain long-term project context across sessions and operate as a proactive engineering partner rather than reactive assistant.

   Problem Statement

   Current conversation memory is session-bound. Each new chat requires re-establishing context about:
   - Project architecture decisions and trade-offs
   - Technology stack specifics (local models, frameworks, deployment targets)
   - Previous iterations and lessons learned
   - Ongoing development challenges

   This creates friction for complex, multi-session projects and forces users to constantly re-explain context.

   Proposed Solution

   Feature 1: Project-Specific Memory Stores

   Description: Allow users to create and manage named "project memory"… more

   3 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  New feature  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

6. Add tools to Lumo so Lumo can gather info from content sources

   Beside the info provided in prompt, Lumo currently has these data sources: Web search, Proton Drive (.txt, .md, .csv, .pdf), and file upload.

   It makes Lumo more powerful if multiple data collecting tools can be added to Lumo. For example, from Proton Mail, from Proton Calendar, from Standard Notes (part of Proton ecosystem).

   Non-Proton sources could be: Google Drive, OneDrive. These can be done by APIs of these services.

   6 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  New feature  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

7. Remove Lumo Completely.

   Please remove lumo completely. It is grift and waste. At MINIMUM provide one button to completely deactivate everywhere. I do not want my subscription funds paying for this. I did not sign up for it, it was added without my consent and i find it very difficult to remove it from my sight.

   3 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   1 comment  ·  Changes to existing features  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

8. Give user direct selection/toggle of Lumo model mode/speed instead of popup menu

   It annoys me that I need to make 4 clicks just bringing back the model type/speed popup to change model from Lite to Max and fast and thinking mode.

   Why aren't there just two toggle switches, one that switches between Lite and Max and another below it that switches between Fats and Thinking. Like this

   Lite <TOGGLE> Max
   Fast <TOGGLE> Thinking

   Put these two toggles right of the input field for the Lumo.

   This way I can quickly switch Lumo model modes in a single or max 2 clicks. Instead of 4 just because I need to bring up the selection popup twice if I want to flip it from Max Thinking to Lite Fast.

   You can put descriptions in a small question mark in a circle next to them just like you explain things in ProtonMail settings.

   It annoys me that I need to make 4 clicks just bringing back the model type/speed popup to change model from Lite to Max and fast and thinking mode.

   Why aren't there just two toggle switches, one that switches between Lite and Max and another below it that switches between Fats and Thinking. Like this

   Lite <TOGGLE> Max
   Fast <TOGGLE> Thinking

   Put these two toggles right of the input field for the Lumo.

   This way I can quickly switch Lumo model modes in a single or max 2 clicks. Instead of 4 just because I need to bring up the… more

   3 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  Changes to existing features  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

9. Integration for Lumo to Home Assistant home automation software

   Add integration to Home Assistant. This would allow Lumo to help with configuration and home automation. Also, its a very tech community with lots of support on YouTube which will help spread the Proton brand further.

   3 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  New feature  ·  Delete…  ·  Admin →
   How important is this to you?

   Not at all You must login first! Important You must login first! Critical You must login first!

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Submit Rating

10. Add secure tool support to Lumo for file creation, previews, and integrations

    Dear Proton Community and Development Team,

    I would like to suggest expanding Lumo with secure tool support so it can become more useful for practical daily work.

    At the moment, an AI assistant is most valuable when it can not only answer questions, but also help users create, organize, preview, and export work in a controlled and privacy-respecting way.

    Suggested tool improvements:

    1. File creation and export
       Lumo should be able to create files such as text documents, tables, simple presentations, ZIP archives, or structured project files and offer them as downloads.

    2. Secure preview features
       Lumo could provide previews for generated content, such as simple web pages, layouts, documents, or structured outputs before the user downloads or applies them.

    3. Live Work inside Lumo
       Lumo should have just like any regular AI, an Canvas feature for example for live editing content or previewing work

    4. Project-style file context
       Users should be able to upload files or reference materials into a Lumo workspace so the assistant can work with them without requiring users to paste very long text manually.

    Why this matters:

    Many users need AI not only for chatting, but also for producing usable outputs. Secure file generation, previews, and optional integrations would make Lumo much more practical while still respecting Proton’s privacy-first approach.

    I understand that this requires careful security and permission handling, but I believe it is possible if implemented with clear user control, transparency, and Proton’s existing privacy standards.

    Thank you for considering this suggestion.

    Dear Proton Community and Development Team,

    I would like to suggest expanding Lumo with secure tool support so it can become more useful for practical daily work.

    At the moment, an AI assistant is most valuable when it can not only answer questions, but also help users create, organize, preview, and export work in a controlled and privacy-respecting way.

    Suggested tool improvements:

    1. File creation and export
       Lumo should be able to create files such as text documents, tables, simple presentations, ZIP archives, or structured project files and offer them as downloads.

    2. Secure preview features
       Lumo could provide previews for generated… more

    5 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

11. Use better open-source ai models.

    Dear Proton Team,

    I would like to propose integrating several high-quality open-source AI models into the Lumo platform. These models align with Proton’s core values of privacy, transparency, and user data sovereignty, while offering excellent performance and flexibility.

    The recommended models include GPT-OSS 120B, LLaMA 3, DeepSeek-V3, Mistral 8x22B, and BLOOM. These models stand out because they are open source, enabling complete self-hosting and customization. This means Proton can maintain full control over user data, keeping all processing within secured European data centers, aligned with Proton’s zero-access policy and strong privacy commitments.

    GPT-OSS 120B offers powerful capabilities close to proprietary large models but without API dependency and high ongoing costs. LLaMA 3 is dialog-optimized and supports long context windows, enhancing user interaction experience. DeepSeek-V3 presents a highly efficient mixture-of-experts architecture with exceptional scaling potential. Mistral 8x22B balances performance and resource demands, fitting well with Proton’s infrastructure. BLOOM provides multi-language support and responsible open-source licensing, broadening Lumo’s global reach.

    By choosing these models, Proton can deliver a privacy-first, performant AI assistant that scales from edge devices to enterprise deployments, all under Proton’s full control. This approach boosts user trust, reduces dependency on third parties, and aligns with evolving global privacy regulations.

    Thank you for considering this proposal.

    Dear Proton Team,

    I would like to propose integrating several high-quality open-source AI models into the Lumo platform. These models align with Proton’s core values of privacy, transparency, and user data sovereignty, while offering excellent performance and flexibility.

    The recommended models include GPT-OSS 120B, LLaMA 3, DeepSeek-V3, Mistral 8x22B, and BLOOM. These models stand out because they are open source, enabling complete self-hosting and customization. This means Proton can maintain full control over user data, keeping all processing within secured European data centers, aligned with Proton’s zero-access policy and strong privacy commitments.

    GPT-OSS 120B offers powerful capabilities close to proprietary… more

    24 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    2 comments  ·  Changes to existing features  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

12. Project Threads Lack Clear Memory Architecture

    Title: Project Threads Lack Clear Memory Architecture—Choose Privacy OR Context, Not Neither
    Category: General Ideas / Lumo

    The Current Problem:

    Lumo's v2.0 persistent memory aggregates content from general chat threads—but excludes project threads from memory aggregation. However, project threads remain indexed and searchable by keyword within the account.

    This creates a contradictory state with no tangible benefit:

    No context continuity — When you mention a project topic in a general chat, Lumo doesn't know the project exists or have any history to reference
    No real privacy — Project content is still searchable/indexed, so it's not a true "black box"
    No security gain — A snooper with account access can find project content via search; a locked memory system wouldn't add protection
    

    Current State:

    Projects: NOT aggregated → memory = incomplete
    Projects: Still indexed → search = works
    Result: Worst of both worlds (loss of convenience, no privacy benefit)
    

    Proposed Solutions (Either/Or):

    Option A: True Private Project Mode

    Project threads are fully gated—requires PIN/code entry to access
    Content excluded from both memory and search indexing
    Opt-in only (user explicitly locks a project)
    For sensitive work that needs genuine isolation
    

    Option B: Full Memory Aggregation

    Project threads feed into Lumo's persistent memory like general chats
    Content indexed and searchable (current behavior)
    Cross-thread context works seamlessly
    For users who prioritize functionality over isolation
    

    Option C: Hybrid Approach (Recommended)

    Default: Project threads aggregate to memory AND indexed
    User can enable "PIN Lock" on specific projects
    Locked projects: Excluded from memory, excluded from search
    Requires authentication to view or discuss locked project content
    Dual-memory system: General memory + Project-only memory (accessible after PIN unlock)
    

    Why This Matters:

    Users managing multiple projects (business startup, technical documentation, etc.) need consistency. If project threads are meant to be private, make them private. If they're meant to support context continuity, let them aggregate.

    The current halfway measure provides neither benefit.

    Requested Action: Implement gatekeeping (PIN/code) for project threads with dedicated project memory upon unlock. Give users explicit control over whether their project content feeds into general memory or stays isolated.

    Title: Project Threads Lack Clear Memory Architecture—Choose Privacy OR Context, Not Neither
    Category: General Ideas / Lumo

    The Current Problem:

    Lumo's v2.0 persistent memory aggregates content from general chat threads—but excludes project threads from memory aggregation. However, project threads remain indexed and searchable by keyword within the account.

    This creates a contradictory state with no tangible benefit:

    No context continuity — When you mention a project topic in a general chat, Lumo doesn't know the project exists or have any history to reference
    No real privacy — Project content is still searchable/indexed, so it's not a true "black box"
    No

    … more
    4 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  Changes to existing features  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

13. Enhance Lumo with Dedicated First-Level Support Training & Automatic Knowledge Updates

    Perhaps it's a generally silly idea, but I often try Lumo as the first point of contact for support regarding Proton products. While the AI assistant delivers good to excellent results in many areas of knowledge, it surprisingly seems to be weak in this area, and I have already had to contend with some misinformation. Therefore, I propose extending Lumo’s capabilities to serve as a reliable, up-to-date first-line support agent. The core idea is to provide Lumo with a focused training set, encompassing, among other things, Proton-specific FAQs, troubleshooting guides, and policy documents, combined with an automated pipeline that feeds the latest developer releases and support team updates directly into the model. Why it matters: Users get instant answers without waiting for a ticket, and routine queries are handled by Lumo, freeing the live support team to focus on complex or security-critical cases.

    Perhaps it's a generally silly idea, but I often try Lumo as the first point of contact for support regarding Proton products. While the AI assistant delivers good to excellent results in many areas of knowledge, it surprisingly seems to be weak in this area, and I have already had to contend with some misinformation. Therefore, I propose extending Lumo’s capabilities to serve as a reliable, up-to-date first-line support agent. The core idea is to provide Lumo with a focused training set, encompassing, among other things, Proton-specific FAQs, troubleshooting guides, and policy documents, combined with an automated pipeline that feeds… more

    6 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    10 comments  ·  Changes to existing features  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

14. Shared Calendar Colors

    So when you share a calendar or invite someone else to it, it would be great or hugely beneficial if it would also save and use the specific colors people have already set for certain event types.
    Because currently it only shows all events in that shared calendar with the default color and the new users would have to change the colors for the respective event individually and manually, which kind of defeats the purpose of a shared environment if you ask me. I mean should each individual user still be able to change colors independently? Sure why not, but I think the default should be: Sync all events and their corresponding colors.

    So when you share a calendar or invite someone else to it, it would be great or hugely beneficial if it would also save and use the specific colors people have already set for certain event types.
    Because currently it only shows all events in that shared calendar with the default color and the new users would have to change the colors for the respective event individually and manually, which kind of defeats the purpose of a shared environment if you ask me. I mean should each individual user still be able to change colors independently? Sure why not, but… more

    4 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  Changes to existing features  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

15. Offline access to Lumo chats & local search (like Proton Drive)

    I'd like to propose a new feature for Lumo that enables access to chat histories without an internet connection – similar to the "Make available offline" function in Proton Drive.

    Use Case: Users want to read and search through specific chats or project histories while offline (e.g., on a plane, in areas with poor connectivity, or for privacy reasons). Currently, Lumo is only usable online, which limits access to your own conversation data.

    Proposed Features:

    1. Offline toggle for chats:
       Like in Proton Drive, users should be able to mark individual chats (or entire projects) with a toggle as "Available offline." These would then be stored locally on the device, fully encrypted.

    2. Two export modes:

    • Full-text mode: The complete chat history is made readable offline (including all details and discussion threads).
    • Summary mode: Users can select multiple chats to generate a compact, Lumo-created summary available offline.

    1. Local search with navigation:

    A full search within the offline-saved chats that:

    • Highlights found terms in color.
    • Allows navigating up/down between search results.
    • Works entirely without internet (indexing happens locally).

    Why this is valuable:

    • Increases the practical usability of Lumo in situations without internet.
    • Provides continuity in long-term projects and conversations, even offline.
    • Leverages proven Proton principles (encrypted local storage) and extends them to chat data.
    • Universally useful – not limited to specific use cases but beneficial for anyone who wants to work flexibly.

    Technical Feasibility: The infrastructure for encrypted local storage already exists in Proton Drive. Extending this to chat data would be a logical evolution. Local search is technically achievable since the text resides on the device after download.

    I'd like to propose a new feature for Lumo that enables access to chat histories without an internet connection – similar to the "Make available offline" function in Proton Drive.

    Use Case: Users want to read and search through specific chats or project histories while offline (e.g., on a plane, in areas with poor connectivity, or for privacy reasons). Currently, Lumo is only usable online, which limits access to your own conversation data.

    Proposed Features:

    1. Offline toggle for chats:
       Like in Proton Drive, users should be able to mark individual chats (or entire projects) with a toggle as "Available offline."… more

    4 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

16. Lumo Conversation Bookmarks

    I often have long conversations with Lumo and weeks or months later need to refer to just a section of that conversation. I know there is a search feature, but I'd like to be able to place a bookmark at a specific point of a conversation instead of conducting a keyword search across all conversations.

    For example, I'm using Lumo for development and had some ideas for future features that I can't get to right now, and I need to keep the conversation on track. Six months from now, I'd like to quickly jump back to the part of the conversation that discussed these features.

    I often have long conversations with Lumo and weeks or months later need to refer to just a section of that conversation. I know there is a search feature, but I'd like to be able to place a bookmark at a specific point of a conversation instead of conducting a keyword search across all conversations.

    For example, I'm using Lumo for development and had some ideas for future features that I can't get to right now, and I need to keep the conversation on track. Six months from now, I'd like to quickly jump back to the part of the… more

    4 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

17. Proton Draw

    Title: Feature Request: Lumo Draw – Privacy-First AI Illustration for Authors (Watercolor & Dark Fantasy Styles)

    Description: I am writing to strongly request the development of Lumo Draw, an AI image generation tool integrated into the Lumo ecosystem with zero-access encryption. (Note: The name "Lumo Draw" was my original concept, proposed here to maintain brand consistency with the Proton ecosystem.)

    My Background & Member History: I have been a loyal Proton member for many years. Currently, I am subscribed to the Lumo+ Family Plan, and I am also a Lifetime Supporter who donated during the annual ticket drive to secure my lifetime status. I am deeply invested in the Proton mission and the privacy-first ecosystem.

    My Use Case as an Author: As an author, I have two distinct illustration needs that current public AI tools cannot safely meet due to data training risks and aggressive censorship:

    Children's Books: I need a dedicated "Warm Watercolor" style. Soft edges, bleeding colors, and whimsical tones are essential. Current tools often force a generic digital look or struggle with consistent style locking.
    Adult Dark Fantasy Novel: This project requires a "Gritty/Dark" aesthetic. It includes atmospheric violence, complex moral themes, and mature storytelling. While it contains no illegal content (CSAM), existing consumer AI tools often over-censor artistic depictions of war, darkness, or mythological figures, making them useless for serious fiction.
    Why This Matters & Why "Lumo Draw"? The market has AI art generators, but they all suffer from the same fatal flaw for creators: They harvest user prompts and images to train future models. If I use Midjourney or DALL-E to design my book characters, I risk losing ownership of my unique IP. My specific "watercolor" or "gritty" style could be ingested into their models and sold back to others.

    A tool named Lumo Draw, operating under Proton's strict zero-access encryption, would ensure my character designs and unpublished story visuals remain mine and are never used to train public models.

    Proposed Features:

    Style Presets: Dedicated modes for "Soft Watercolor" (warmth/fantasy) and "Dark/Gritty" (high contrast/noir).
    Privacy First: No logs, no scanning of creative work for "policy violations" unless illegal, ensuring my artistic vision isn't sanitized by corporate filters.
    Consistency: Tools to lock character references across multiple generated pages.
    Paid Tier Willingness: As a long-time member and Lifetime supporter, I am explicitly willing to increase my payment to access a premium "Lumo Draw" service tier.
    Conclusion: Lumo Draw would be a game-changer for independent authors who value both creativity and privacy. It would allow us to visualize our worlds without compromising our intellectual property. Having suggested this name and concept after years of supporting Proton, I believe this feature aligns perfectly with the community's needs.

    Thank you for considering this feedback from a dedicated long-term member.

    Title: Feature Request: Lumo Draw – Privacy-First AI Illustration for Authors (Watercolor & Dark Fantasy Styles)

    Description: I am writing to strongly request the development of Lumo Draw, an AI image generation tool integrated into the Lumo ecosystem with zero-access encryption. (Note: The name "Lumo Draw" was my original concept, proposed here to maintain brand consistency with the Proton ecosystem.)

    My Background & Member History: I have been a loyal Proton member for many years. Currently, I am subscribed to the Lumo+ Family Plan, and I am also a Lifetime Supporter who donated during the annual ticket drive to secure… more

    3 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    2 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

18. Allow Lumo to deal with Memory reduplicates

    I notice the memory list often has either exact duplicates or very similar worded info. I presume there is some kind of self-check to see if a memory has already been added. But its not quite up to standard.
    May i suggest can we give Lumo permission to comb through the memory and weed out duplicates or too similar ones. If there are edge cases it could apply a tag for a special filter called User Attention Needed (or something) and we can click on this button that is lit up when it has memories to assess and we only see the list of edge case memories and each one has a green tick button to approve or X to delete and a pencil icon to edit and once done we can tap the green tick button to allow it back into the main memories (and of course remove the special filter tag)

    Also to further enhance this idea, add search bar so we can see if what certain topics might be listed. And we could also filter our own manual memory entries with a command in the search bar: #you or /you

    I notice the memory list often has either exact duplicates or very similar worded info. I presume there is some kind of self-check to see if a memory has already been added. But its not quite up to standard.
    May i suggest can we give Lumo permission to comb through the memory and weed out duplicates or too similar ones. If there are edge cases it could apply a tag for a special filter called User Attention Needed (or something) and we can click on this button that is lit up when it has memories to assess and we only… more

    4 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

19. Export and Access Chat-Knowledge Files to/in Proton Drive

    Perhaps, a 1-click button to export chat-knowledge files and/or transcripts (with formatting & timestamps reserved) to Proton Drive would be nice? And/or, having Lumo access files in Proton Drive as chat-knowledge files would also be useful, especially given the file-upload limit per chat.

    6 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    0 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating

20. To make Lumo more human

    Subject: Proposal: A "Dual-Loop" Architecture for Next-Gen LLMs Inspired by Neuro-Mechanical Models

    Dear Research and Engineering Team,

    I am writing to share a conceptual proposal for a new architectural approach to Large Language Models (LLMs), inspired by the neuro-mechanical theories of Lee Kent Hempfling (specifically the distinction between Short-Term and Long-Term memory loops).

    While current LLMs excel at pattern recognition and statistical prediction (effectively acting as massive Long-Term Memory systems), they lack a mechanism for inhibition, self-correction, and proactive intent. This proposal suggests a "Dual-Loop" architecture that could significantly reduce hallucinations, bias, and reactive behaviour.
    The Core Concept: Dual-Loop Architecture

    Current models rely on a single stream of attention mechanisms. I propose a system with two distinct, interacting loops:

    The Long-Term Loop (The Library):
        Function: Standard Transformer weights for retrieving facts, grammar, and historical patterns.
        Role: Fast, reactive, and high-volume data retrieval.
    The Short-Term Loop (The Operator):
        Function: A separate, high-speed module running in parallel.
        Role: Inhibition and Monitoring. Before the Long-Term loop commits to a token, the Short-Term loop evaluates:
            Is this consistent with the immediate context?
            Is this a hallucination or a bias?
            Does this align with the user's intent?
        Mechanism: If a conflict is detected, the Short-Term loop inhibits the output and triggers a re-evaluation cycle (simulating the human "brake" or "pause").
    

    Key Proposed Features

    Dynamic Inhibition Layers: Instead of just predicting the next token, the model actively suppresses low-probability or biased outputs that fail a "truth/intent" check.
    Frequency-Based Temporal Processing: Implementing a "Biological Clock" simulation where different parts of the network process information at different speeds (Fast for "Now"/Logic, Slow for Deep Memory), creating a simulated sense of time and continuity.
    Intent-Driven Generation: Shifting from "Pattern Completion" to "Goal Achievement," where the Short-Term loop sets a constraint/goal before the Long-Term loop retrieves data.
    Modality-Aware Routing: Detecting if a query requires visual simulation or logical deduction and routing processing accordingly, similar to how humans switch between visual and aural dominance.
    

    Potential Benefits

    Reduced Hallucinations: By inhibiting outputs that don't pass a consistency check.
    Lower Bias: Active suppression of training-data biases via the Short-Term "brake."
    Improved Alignment: Better adherence to user intent rather than just statistical likelihood.
    Organic Output: Moving away from rigid, "stiff" text generation toward fluid, "pro-active" responses.
    

    Next Steps

    I believe this approach moves beyond simple scaling (adding more parameters) to architectural innovation. It treats the LLM not just as a database, but as a system capable of self-regulation.

    I would be happy to provide further details, diagrams, or specific mathematical formulations for the "Inhibition Layer" if the team finds this direction promising.

    This Idea was created by Lumo with prompts from me. I thought it was good enough for others to have a better look and did'nt know who to send it to.

    Subject: Proposal: A "Dual-Loop" Architecture for Next-Gen LLMs Inspired by Neuro-Mechanical Models

    Dear Research and Engineering Team,

    I am writing to share a conceptual proposal for a new architectural approach to Large Language Models (LLMs), inspired by the neuro-mechanical theories of Lee Kent Hempfling (specifically the distinction between Short-Term and Long-Term memory loops).

    While current LLMs excel at pattern recognition and statistical prediction (effectively acting as massive Long-Term Memory systems), they lack a mechanism for inhibition, self-correction, and proactive intent. This proposal suggests a "Dual-Loop" architecture that could significantly reduce hallucinations, bias, and reactive behaviour.
    The Core Concept: Dual-Loop Architecture

    … more

    4 votes

    Vote Vote Vote Vote

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Vote

    We’ll send you updates on this idea

    3 comments  ·  New feature  ·  Delete…  ·  Admin →
    How important is this to you?

    Not at all You must login first! Important You must login first! Critical You must login first!

    We're glad you're here

    Please sign in to leave feedback

    Signed in as (Sign out)

    Close

    Close

    Submit Rating