docs: document per-agent API keys and rate limiting in .env.example

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: Skiipy11

.env.example

@@ -1,10 +1,24 @@
 # Shared Brain Configuration
 
 # --- Required ---
-BRAIN_API_KEY=your-api-key-here
+BRAIN_API_KEY=your-admin-api-key-here
 PORT=8084
 # HOST=0.0.0.0  # Default: 127.0.0.1 (localhost only). Set to 0.0.0.0 for LAN/Docker access.
 
+# --- Per-Agent API Keys (optional, recommended) ---
+# Each agent gets its own key. The key name maps to the agent identity:
+# AGENT_KEY_<name> → agent authenticates as "<name>" (underscores become hyphens)
+# When using an agent key, source_agent must match the authenticated identity.
+# BRAIN_API_KEY remains as admin/fallback (no identity binding).
+# Generate keys with: openssl rand -hex 32
+# AGENT_KEY_claude_code=your-key-here
+# AGENT_KEY_n8n=your-key-here
+# AGENT_KEY_morpheus=your-key-here
+
+# --- Rate Limiting ---
+# RATE_LIMIT_WRITES=60   # Max write requests per minute per key (default: 60)
+# RATE_LIMIT_READS=120   # Max read requests per minute per key (default: 120)
+
 # --- Qdrant (vector store — always required) ---
 QDRANT_URL=http://qdrant:6333
 QDRANT_API_KEY=your-qdrant-key-here