brain-mcp

Instructions

    Reconstruct cognitive save-state for a domain — where you left off.
    Returns: thinking stage, open questions, decisions, concepts, emotional tone.
    The 'load game' button for your mind.
    

Implementation Reference

• brain_mcp/server/tools_prosthetic.py:108-205 (handler)

  The 'tunnel_state' tool implementation in brain_mcp/server/tools_prosthetic.py. It reconstructs the cognitive save-state for a given domain by querying the summaries database or falling back to raw conversation data.

  def tunnel_state(domain: str, limit: int = 10) -> str:
      """
      Reconstruct cognitive save-state for a domain — where you left off.
      Returns: thinking stage, open questions, decisions, concepts, emotional tone.
      The 'load game' button for your mind.
      """
      db = get_summaries_db()
      if db:
          rows = db.execute("""
              SELECT summary, thinking_stage, importance, emotional_tone,
                     open_questions, decisions, concepts, key_insights, connections_to,
                     cognitive_pattern, problem_solving_approach, msg_count, title, source
              FROM summaries
              WHERE domain_primary = ?
              ORDER BY summarized_at DESC
              LIMIT ?
          """, [domain, limit]).fetchall()
  
          if not rows:
              return f"No conversations found for domain: {domain}"
  
          cols = [
              "summary", "thinking_stage", "importance", "emotional_tone",
              "open_questions", "decisions", "concepts", "key_insights",
              "connections_to", "cognitive_pattern", "problem_solving_approach",
              "msg_count", "title", "source",
          ]
  
          latest = dict(zip(cols, rows[0]))
  
          all_oq, all_dec, all_concepts, all_insights, all_connections = [], [], set(), [], set()
          importance_counts = {"breakthrough": 0, "significant": 0, "routine": 0}
          stage_counts = {}
  
          for row in rows:
              r = dict(zip(cols, row))
              for q in parse_json_field(r["open_questions"]):
                  if q and q.lower() != "none identified" and q not in all_oq:
                      all_oq.append(q)
              for d in parse_json_field(r["decisions"]):
                  if d and d not in all_dec:
                      all_dec.append(d)
              for c in parse_json_field(r["concepts"]):
                  all_concepts.add(c)
              for i in parse_json_field(r["key_insights"]):
                  if i and i not in all_insights:
                      all_insights.append(i)
              for c in parse_json_field(r["connections_to"]):
                  all_connections.add(c)
              imp = r["importance"] or "routine"
              importance_counts[imp] = importance_counts.get(imp, 0) + 1
              stage = r["thinking_stage"] or ""
              stage_counts[stage] = stage_counts.get(stage, 0) + 1
  
          output = [f"## 🧠 Tunnel State: {domain}\n"]
          output.append(f"**Current stage:** {latest['thinking_stage'] or 'unknown'}")
          output.append(f"**Emotional tone:** {latest['emotional_tone'] or 'unknown'}")
          output.append(f"**Conversations:** {len(rows)} (last {limit})")
          bt = importance_counts.get("breakthrough", 0)
          if bt:
              output.append(f"**Breakthroughs:** {bt} 💎")
          output.append(f"**Cognitive pattern:** {latest['cognitive_pattern'] or 'unknown'}")
          output.append(f"**Problem solving:** {latest['problem_solving_approach'] or 'unknown'}")
  
          if all_oq:
              output.append(f"\n### ❓ Open Questions ({len(all_oq)})")
              for q in all_oq[:10]:
                  output.append(f"  - {q}")
              if len(all_oq) > 10:
                  output.append(f"  _... and {len(all_oq)-10} more_")
  
          if all_dec:
              output.append(f"\n### ✅ Decisions ({len(all_dec)})")
              for d in all_dec[:7]:
                  output.append(f"  - {d}")
              if len(all_dec) > 7:
                  output.append(f"  _... and {len(all_dec)-7} more_")
  
          if all_concepts:
              output.append(f"\n### 🏷️ Active Concepts ({len(all_concepts)})")
              output.append(f"  {', '.join(sorted(all_concepts)[:15])}")
  
          if all_insights:
              output.append(f"\n### 💡 Key Insights")
              for i in all_insights[:5]:
                  output.append(f"  - {i}")
  
          if all_connections:
              output.append(f"\n### 🔗 Connected Domains")
              output.append(f"  {', '.join(sorted(all_connections)[:10])}")
  
          if stage_counts:
              output.append(f"\n### 📊 Thinking Stage History")
              for s, c in sorted(stage_counts.items(), key=lambda x: -x[1]):
                  output.append(f"  {s or 'unknown'}: {c}")
  
          return "\n".join(output)