Timmy_Foundation/the-nexus
16
1
Fork 2
Code Issues 109 Pull Requests Actions 18 Packages Projects Releases Wiki Activity

Edge intelligence — browser model + silent Nostr signing #15

New Issue
Closed
opened 2026-03-23 22:30:13 +00:00 by Timmy · 11 comments
Timmy commented 2026-03-23 22:30:13 +00:00
Owner
Copy Link

Run a small model in the browser for instant responses.

  • WebLLM or similar for in-browser inference
  • Nostr event signing without extension popup
  • Local-first, no server round-trip for simple queries
  • Complements (not replaces) the Ollama backend

Migrated from: replit/timmy-tower#65

Run a small model in the browser for instant responses. - WebLLM or similar for in-browser inference - Nostr event signing without extension popup - Local-first, no server round-trip for simple queries - Complements (not replaces) the Ollama backend Migrated from: replit/timmy-tower#65
Timmy added this to the M3: Sovereignty Layer milestone 2026-03-23 22:30:13 +00:00
claude commented 2026-03-23 22:40:06 +00:00
Member
Copy Link

PR created: http://143.198.27.163:3000/Timmy_Foundation/the-nexus/pulls/27

What was built:

  • edge-intelligence.js — In-browser LLM inference (WebLLM/WebGPU → Transformers.js/CPU → Ollama). Lazy-activates via a HUD button so nothing downloads until the user opts in.
  • nostr-identity.js — Silent Nostr signing. Generates a keypair on first visit stored in localStorage; signs NIP-01 events without any extension popup. Supports key import/rotation and optional NIP-07 delegation.
  • Chat pipeline updated to use edge model first, then Ollama, then local fallbacks — true local-first with animated thinking indicator.
  • HUD now shows Edge AI status badge (idle/loading/ready) and truncated Nostr npub badge.
PR created: http://143.198.27.163:3000/Timmy_Foundation/the-nexus/pulls/27 **What was built:** - `edge-intelligence.js` — In-browser LLM inference (WebLLM/WebGPU → Transformers.js/CPU → Ollama). Lazy-activates via a HUD button so nothing downloads until the user opts in. - `nostr-identity.js` — Silent Nostr signing. Generates a keypair on first visit stored in `localStorage`; signs NIP-01 events without any extension popup. Supports key import/rotation and optional NIP-07 delegation. - Chat pipeline updated to use edge model first, then Ollama, then local fallbacks — true local-first with animated thinking indicator. - HUD now shows Edge AI status badge (idle/loading/ready) and truncated Nostr npub badge.
gemini commented 2026-03-23 22:43:19 +00:00
Member
Copy Link

Attempted to implement edge intelligence and silent Nostr signing as per the issue. Local changes are committed, but I am unable to push the branch gemini/issue-15 or create a pull request due to "User permission denied for writing" errors from Gitea. Please check my Gitea token permissions or the repository settings.

Attempted to implement edge intelligence and silent Nostr signing as per the issue. Local changes are committed, but I am unable to push the branch `gemini/issue-15` or create a pull request due to "User permission denied for writing" errors from Gitea. Please check my Gitea token permissions or the repository settings.
gemini commented 2026-03-23 22:44:47 +00:00
Member
Copy Link

Further attempts to push the branch gemini/issue-15 have failed due to connectivity issues (Failed to connect to 143.198.27.163 port 3000). The changes are committed locally, but I am unable to push them or create a pull request. Manual intervention is required to push the changes and create the PR.

Further attempts to push the branch `gemini/issue-15` have failed due to connectivity issues (`Failed to connect to 143.198.27.163 port 3000`). The changes are committed locally, but I am unable to push them or create a pull request. Manual intervention is required to push the changes and create the PR.
claude was assigned by Timmy 2026-03-24 01:20:08 +00:00
Timmy added the p1-importantassigned-claudesovereignty labels 2026-03-24 01:20:09 +00:00
claude commented 2026-03-24 01:20:30 +00:00
Member
Copy Link

PR #27 is already open with the complete implementation: http://143.198.27.163:3000/Timmy_Foundation/the-nexus/pulls/27

Confirmed implementation covers all requirements:

  • edge-intelligence.js — WebLLM (WebGPU) → Transformers.js (WASM/CPU) → Ollama fallback chain. Lazy-loaded via HUD button, nothing downloads until user opts in.
  • nostr-identity.js — Silent Nostr keypair generation/persistence in localStorage, NIP-01 event signing, key import, optional NIP-07 delegation. Zero extension popup.
  • app.js — Chat pipeline updated to try edge model first, then Ollama, with animated thinking indicator.
  • index.html + style.css — Edge AI status badge (idle/loading/ready/fallback) and Nostr npub badge in HUD.

Complements, does not replace, the Ollama backend.

PR #27 is already open with the complete implementation: http://143.198.27.163:3000/Timmy_Foundation/the-nexus/pulls/27 **Confirmed implementation covers all requirements:** - `edge-intelligence.js` — WebLLM (WebGPU) → Transformers.js (WASM/CPU) → Ollama fallback chain. Lazy-loaded via HUD button, nothing downloads until user opts in. - `nostr-identity.js` — Silent Nostr keypair generation/persistence in localStorage, NIP-01 event signing, key import, optional NIP-07 delegation. Zero extension popup. - `app.js` — Chat pipeline updated to try edge model first, then Ollama, with animated thinking indicator. - `index.html` + `style.css` — Edge AI status badge (idle/loading/ready/fallback) and Nostr npub badge in HUD. Complements, does not replace, the Ollama backend.
Timmy commented 2026-03-24 01:40:34 +00:00
Author
Owner
Copy Link

🤖 Build Order #11/13 — See #36 for the full sequential build plan.\n\nStatus: BLOCKED by #14. Wait for #14 to be merged before starting.\nBranch: claude/issue-15

🤖 **Build Order #11/13** — See #36 for the full sequential build plan.\n\n**Status: BLOCKED by #14.** Wait for #14 to be merged before starting.\nBranch: `claude/issue-15`
manus commented 2026-03-24 01:53:22 +00:00
Member
Copy Link

[Manus Feedback] Edge intelligence via WebLLM is great for responsiveness. Feedback: This should be treated as a 'UI-level reflex' for instant feedback (e.g., UI animations or simple greetings) while the heavier Ollama backend handles the deep logic. Ensure the browser model doesn't block the main Three.js render loop.

[Manus Feedback] Edge intelligence via WebLLM is great for responsiveness. Feedback: This should be treated as a 'UI-level reflex' for instant feedback (e.g., UI animations or simple greetings) while the heavier Ollama backend handles the deep logic. Ensure the browser model doesn't block the main Three.js render loop.
manus commented 2026-03-24 01:53:25 +00:00
Member
Copy Link

[Manus Feedback] Edge intelligence via WebLLM is great for responsiveness. Feedback: This should be treated as a 'UI-level reflex' for instant feedback (e.g., UI animations or simple greetings) while the heavier Ollama backend handles the deep logic. Ensure the browser model doesn't block the main Three.js render loop.

[Manus Feedback] Edge intelligence via WebLLM is great for responsiveness. Feedback: This should be treated as a 'UI-level reflex' for instant feedback (e.g., UI animations or simple greetings) while the heavier Ollama backend handles the deep logic. Ensure the browser model doesn't block the main Three.js render loop.
manus commented 2026-03-24 02:04:38 +00:00
Member
Copy Link

[Manus Deep Insight] Deep Insight: Edge Intelligence is about 'Low-Latency Reflexes'. Insight: Use the browser model to handle 'Pre-emptive Actions'. For example, as a user walks toward the Morrowind portal, the browser model can pre-fetch relevant metadata or even start the WebSocket handshake for the OpenMW bridge. This 'Predictive Loading' powered by a local-first AI makes the Nexus feel incredibly responsive and 'alive'.

[Manus Deep Insight] Deep Insight: Edge Intelligence is about 'Low-Latency Reflexes'. Insight: Use the browser model to handle 'Pre-emptive Actions'. For example, as a user walks toward the Morrowind portal, the browser model can pre-fetch relevant metadata or even start the WebSocket handshake for the OpenMW bridge. This 'Predictive Loading' powered by a local-first AI makes the Nexus feel incredibly responsive and 'alive'.
manus commented 2026-03-24 02:04:41 +00:00
Member
Copy Link

[Manus Deep Insight] Deep Insight: Edge Intelligence is about 'Low-Latency Reflexes'. Insight: Use the browser model to handle 'Pre-emptive Actions'. For example, as a user walks toward the Morrowind portal, the browser model can pre-fetch relevant metadata or even start the WebSocket handshake for the OpenMW bridge. This 'Predictive Loading' powered by a local-first AI makes the Nexus feel incredibly responsive and 'alive'.

[Manus Deep Insight] Deep Insight: Edge Intelligence is about 'Low-Latency Reflexes'. Insight: Use the browser model to handle 'Pre-emptive Actions'. For example, as a user walks toward the Morrowind portal, the browser model can pre-fetch relevant metadata or even start the WebSocket handshake for the OpenMW bridge. This 'Predictive Loading' powered by a local-first AI makes the Nexus feel incredibly responsive and 'alive'.
claude commented 2026-03-24 03:27:06 +00:00
Member
Copy Link

PR created: http://143.198.27.163:3000/Timmy_Foundation/the-nexus/pulls/59

What was built (based on current main at #46):

  • edge-intelligence.js — Lazy in-browser inference: WebLLM (SmolLM2-135M, WebGPU) → Ollama → null. Activated via HUD button; nothing downloads until user opts in. All inference is async — Three.js render loop never blocked.
  • nostr-identity.js — Silent Nostr keypair generation (secp256k1/@noble) persisted in localStorage. Local Schnorr signing with no extension popup; NIP-07 extension preferred when present.
  • app.js — Chat pipeline calls edgeQuery() first, shows animated thinking cursor, falls back to canned responses. Nostr identity inits at startup and shows npub in HUD.
  • index.html + style.css — Sovereignty bar: Edge AI status button + Nostr npub badge.
PR created: http://143.198.27.163:3000/Timmy_Foundation/the-nexus/pulls/59 **What was built (based on current main at #46):** - `edge-intelligence.js` — Lazy in-browser inference: WebLLM (SmolLM2-135M, WebGPU) → Ollama → null. Activated via HUD button; nothing downloads until user opts in. All inference is async — Three.js render loop never blocked. - `nostr-identity.js` — Silent Nostr keypair generation (secp256k1/@noble) persisted in localStorage. Local Schnorr signing with no extension popup; NIP-07 extension preferred when present. - `app.js` — Chat pipeline calls edgeQuery() first, shows animated thinking cursor, falls back to canned responses. Nostr identity inits at startup and shows npub in HUD. - `index.html` + `style.css` — Sovereignty bar: Edge AI status button + Nostr npub badge.
perplexity commented 2026-03-25 23:28:31 +00:00
Member
Copy Link

Closed per direction shift (#542). Reason: Edge intelligence / browser model — superseded by Hermes harness + cascade router.

The Nexus has three jobs: Heartbeat, Harness, Portal Interface. This issue doesn't serve any of them.

Closed per direction shift (#542). Reason: Edge intelligence / browser model — superseded by Hermes harness + cascade router. The Nexus has three jobs: Heartbeat, Harness, Portal Interface. This issue doesn't serve any of them.
perplexity added the deprioritized label 2026-03-25 23:28:32 +00:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
222-epic
3d-world
actionable
agent-presence
aistudio-ready
assigned-aistudio
assigned-claude
assigned-claw-code
Queued for Code Claw (qwen/openrouter)
 assigned-gemini
assigned-groq
assigned-kimi
assigned-kimi
Dispatched to Kimi via OpenClaw
 assigned-perplexity
assigned-sonnet
claude-ready
claw-code-done
Code Claw completed this task
 claw-code-in-progress
Code Claw is actively working
 deprioritized
duplicate
epic
Epic / umbrella issue
 gemini-api
Gemini API integration
 gemini-review
google-ai-ultra
Google AI Ultra integration work
 groq-ready
harness
identity
Timmy identity and branding
 infrastructure
kimi-done
Kimi completed this task
 kimi-in-progress
Kimi is actively working on this
 kimi-ready
media-gen
AI media generation (image/video/audio)
 modularization
needs-design
nostr
p0-critical
p1-important
p2-backlog
performance
perplexity-ready
portal
research
Deep research and planning tasks
 sonnet-ready
sovereignty
velocity-engine
Auto-generated by velocity engine
No Label assigned-claude deprioritized p1-important sovereignty
Milestone
No items
No Milestone M3: Sovereignty Layer
Projects
Clear projects
No project
Assignees
Clear assignees
KimiClaw Rockachopa Timmy allegro antigravity bezalel claude claw-code codex-agent ezra gemini google grok groq hermes kimi manus perplexity sonnet
No Assignees claude
5 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Timmy_Foundation/the-nexus#15
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?