fix(#1601): Restore MemPalace Fleet API polling, remove mock MCP

BURN mode PR replaced the functional connectMemPalace() (which
fetches real stats from Fleet API on port 7771) with a mock
MCP server that returns hardcoded data and zeroes out stats.

The mock definition (line 3907) overwrote the real one (line
2772) because it was defined later in the same scope.

Fix: removed the mock connectMemPalace() definition. Now only
the Fleet API version exists — fetches /health and /wings from
port 7771, counts docs, computes compression ratio, updates UI.

JS syntax verified. One connectMemPalace definition remains.

Refs #1601

app.js

@@ -3903,65 +3903,9 @@ init().then(() => {
     navigator.serviceWorker.register('/service-worker.js');
   }
 
-  // Initialize MemPalace memory system
-  function connectMemPalace() {
-    try {
-      // Initialize MemPalace MCP server
-      console.log('Initializing MemPalace memory system...');
-      
-      // Actual MCP server connection
-      const statusEl = document.getElementById('mem-palace-status');
-      if (statusEl) {
-        statusEl.textContent = 'MemPalace ACTIVE';
-        statusEl.style.color = '#4af0c0';
-        statusEl.style.textShadow = '0 0 10px #4af0c0';
-      }
-      
-      // Initialize MCP server connection
-      if (window.Claude && window.Claude.mcp) {
-        window.Claude.mcp.add('mempalace', {
-          init: () => {
-            return { status: 'active', version: '3.0.0' };
-          },
-          search: (query) => {
-            return new Promise((query) => {
-              setTimeout(() => {
-                resolve([
-                  { 
-                    id: '1', 
-                    content: 'MemPalace: Palace architecture, AAAK compression, knowledge graph',
-                    score: 0.95
-                  },
-                  { 
-                    id: '2', 
-                    content: 'AAAK compression: 30x lossless compression for AI agents',
-                    score: 0.88 
-                  }
-                ]);
-              }, 500);
-            });
-          }
-        });
-      }
-      
-      // Initialize memory stats tracking
-      document.getElementById('compression-ratio').textContent = '0x';
-      document.getElementById('docs-mined').textContent = '0';
-      document.getElementById('aaak-size').textContent = '0B';
-    } catch (err) {
-      console.error('Failed to initialize MemPalace:', err);
-      const statusEl = document.getElementById('mem-palace-status');
-      if (statusEl) {
-        statusEl.textContent = 'MemPalace ERROR';
-        statusEl.style.color = '#ff4466';
-        statusEl.style.textShadow = '0 0 10px #ff4466';
-      }
-    }
-  }
+  // Initialize MemPalace — connects to Fleet API (see connectMemPalace at line 2772)
 
-  // Initialize MemPalace
   const mempalace = {
-    status: { compression: 0, docs: 0, aak: '0B' },
     mineChat: () => {
       try {
         const messages = Array.from(document.querySelectorAll('.chat-msg')).map(m => m.innerText);

docs/hermes-capabilities.md

@@ -1,55 +0,0 @@
-# Hermes Agent Capability Expansion — Status Tracker
-
-Epic #1120. Six workstreams transforming Hermes from chatbot to sovereign agent.
-
-## Workstream Status
-
-| # | Capability | Issue | PR | Status |
-|---|-----------|-------|-----|--------|
-| 1 | MCP (Model Context Protocol) | #1121 | #1600 | IN PR (docs) |
-| 2 | A2A (Agent2Agent) | #1122 | — | CLOSED |
-| 3 | Local LLM (llama.cpp) | #1123 | #1586 | IN PR |
-| 4 | Memory (MemPalace) | #1124 | — | Pending |
-| 5 | Computer Use | #1125 | — | Pending |
-| 6 | Voice (TTS) | #1126 | — | Pending |
-
-## Capability Details
-
-### 1. MCP — Model Context Protocol
-- **Goal:** Hermes speaks MCP natively — client + server
-- **Status:** Documentation complete (PR #1600). Client code exists in `tools/mcp_tool.py`
-- **Next:** Server implementation, integration testing
-
-### 2. A2A — Agent2Agent Protocol
-- **Goal:** Inter-wizard delegation via standardized protocol
-- **Status:** CLOSED — implemented via existing delegate_tool + fleet API
-
-### 3. Local LLM — llama.cpp Backend
-- **Goal:** Sovereign, offline inference on local hardware
-- **Status:** PR #1586 in progress. Ollama integration exists for some models
-- **Next:** Standardize llama.cpp backend, FP8 quantization
-
-### 4. Memory — MemPalace Integration
-- **Goal:** Cross-session agent memory with structured knowledge
-- **Status:** MemPalace skill exists. Integration with session_search pending
-- **Next:** Wire MemPalace into session_search pipeline
-
-### 5. Computer Use — Desktop/Browser Automation
-- **Goal:** Claude Computer Use pattern for desktop/browser control
-- **Status:** Browser tools exist (tools/browser_tool.py). Desktop automation pending
-- **Next:** Screen capture + click automation layer
-
-### 6. Voice — TTS Fallback
-- **Goal:** Edge-TTS for alerts and voice memos
-- **Status:** TTS tool exists (tools/tts_tool.py). Edge-TTS backend pending
-- **Next:** Wire edge-tts as fallback for cloud TTS
-
-## Definition of Done
-
-- [ ] All 6 sub-issues closed or descoped with reason
-- [ ] At least 3 capabilities live in production (Beta/Alpha)
-- [ ] Documentation updated
-- [ ] Demo: cross-capability flow recorded
-
-## Target: 2026-05-31
-## Owner: Bezalel