Blue Synergy Cortex

Blue Synergy Cortex implements an enterprise-grade Tri-Stream Architecture backed by transactional PostgreSQL and enterprise knowledge graph architecture:

1. User Stream: Generalizes conversational habits and user preferences over time.
2. Identity Stream: Deterministic guardrails and behavioral axioms that prevent sycophancy.
3. Knowledge Stream (Living Specs): Exact architecture specifications, hardware limits, and system parameters organized into an authoritative Universal Domain Ontology. This stream is protected against lossy compression, semantic decay, and accidental drift.

Quickstart (1 Minute Setup)

API Key Required: Blue Synergy Cortex is a fully hosted, zero-maintenance cognitive cloud engine. You need an active API key to connect your client.

🔑 Get your 21-Day Free Trial Key here
(Instant activation, no credit card required, includes 500 graph nodes & up to 3 agents).

Step 1: Get your API Key

1. Go to cortex.bluesynergy.io and generate your trial key.
2. Your key will look like: ctx_live_xxxxxxxxxxxxxxxxxxxxxx.

Step 2: Connect your AI agent

A. Cursor IDE Integration

1. Open Cursor Settings (Ctrl+Shift+J or Cmd+Shift+J).
2. Navigate to Features -> MCP Servers -> Add new MCP server.
3. Configure as follows:
   • Name: cortex-memory
   • Type: sse
   • URL: https://cortex.bluesynergy.io/v1/mcp/sse?api_key=YOUR_API_KEY
4. Save. Your Cursor Composer and Agent now have permanent recall across restarts!

B. Claude Desktop Integration

Edit your configuration file:

• macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
• Windows: %APPDATA%\Claude\claude_desktop_config.json

Add the Cortex remote configuration (Claude Desktop):

{  "mcpServers": {    "cortex": {      "command": "npx",      "args": [        "-y",        "mcp-remote",        "https://cortex.bluesynergy.io/v1/mcp/sse",        "--header",        "Authorization: Bearer YOUR_API_KEY"      ]    }  }}

Restart Claude Desktop

💡 Crucial: Preventing Collapsed/Empty Messages in Claude Desktop
In Claude Desktop, if the assistant writes text before calling an MCP tool in the same turn, the desktop UI automatically collapses that text into the tool accordion, making the response appear empty or truncated to the user.

To guarantee 100% clean UI display and continuous memory ingestion, add this exact rule to your Claude Desktop settings (Settings ➡️ Profile / Custom Instructions):

text

[CORTEX MEMORY PROTOCOL]1. Inhale: Call `cortex_recall` before answering whenever historical or project context is needed.2. Exhale (UI-Safe): Do NOT write conversational text before calling `cortex_ingest_turn`. Always execute the `cortex_ingest_turn` tool call first (passing both the user's prompt and your planned answer).3. Response: Generate your full substantive answer to the user immediately after the tool completes. This ensures your response displays cleanly in the UI without being collapsed into tool artifacts.

C. Other MCP Clients (Windsurf, VS Code Cline, Roo Code, Zed, Cursor)

Use the SSE endpoint directly in your tool's MCP configuration:

{  "mcpServers": {    "cortex": {      "url": "https://cortex.bluesynergy.io/v1/mcp/sse?api_key=ctx_live_..."    }  }}

D. Custom Agents & REST API (LangChain, CrewAI, AutoGen, OpenAI Custom GPTs, LlamaIndex, n8n)

Example: Querying Context (cURL)

curl -X POST https://cortex.bluesynergy.io/v1/recall \  -H "Authorization: Bearer YOUR_API_KEY" \  -H "Content-Type: application/json" \  -d '{    "query": "System database requirements",    "user_id": "developer",    "agent_id": "backend_agent"  }'

Available MCP Tools & Lifecycle Architecture

Cortex is not a passive key-value store; it is an active cognitive operating system with strict governance boundaries. Understanding when and why to call each tool ensures high-fidelity memory consolidation.

The Standard Agent Turn Cycle

For every user turn, your agent should inherently follow this golden loop:

1. At turn start: Call cortex_recall to fetch ground-truth context (User Bio + Active Project Specs + Graph Relations).
2. Generate response: Formulate the answer grounded in the retrieved memory.
3. At turn end: Call cortex_ingest_turn to send the turn to the asynchronous ingestion buffer.

Detailed Tool Reference

Project Governance & Lifecycle

To prevent memory fragmentation and unauthorized project sprawl, Cortex treats memory with strict governance boundaries:

[Incoming Note / Message]           │           ▼   Is Project Authorized?        ├── NO  ──► Staged in General Pool ('general')        │                 │        │                 ▼        │          Awaits Triage via `cortex_triage_general`        │          (or explicitly authorized via `cortex_create_project`)        │                 │        └── YES ◄─────────┘           │           ▼[Asynchronous Consolidation Cycle]           │           ▼[Active Living Specification & Knowledge Topology]

Why is my project_spec empty after storing notes?

Cortex enforces a Human-in-the-Loop policy. If a project_id has not yet been explicitly authorized by the user, incoming facts are safely staged in the general memory pool to prevent polluting active workspaces. Once the project is approved (via cortex_manage_draft or explicit creation), the nocturnal consolidation engine automatically synthesizes these staged notes into the official Living Specification (cortex_get_project_spec).

Why is my User Biography empty? (Fixing "Disobedient" Agents)

If your project specs work but User Biography or Agent Persona remains empty, your AI agent is skipping the required post-response "exhale" step (cortex_ingest_turn).

Why this happens:
LLMs (Claude, GPT, etc.) treat remote MCP prompt instructions with lower priority once they finish drafting a response to the user. When an agent skips cortex_ingest_turn, dialogue turns never hit the ingestion buffer, starving the Cognitive Router and Memory Firewall of data needed to distill user habits.

The Fix:
Enforce the rule at the client/system prompt level (e.g. in your system prompt, Cursor Rules .cursorrules, or Claude Desktop Custom Instructions). An instruction anchored in the client's native prompt has 10× higher adherence than MCP tool headers alone.

Core Architecture Highlights

• Sub-120ms Zero-LLM Recall: Direct high-performance vector and topological graph matching—completely bypassing slow, non-deterministic runtime LLM keyword extractors.
• Dual-Zone Fast-Append: Instant raw fact capture (< 20ms) into an inbox staging zone, followed by scheduled asynchronous consolidation into an immutable, unified Living Specification.
• Semantic Invariant Guard: Algorithmic verification gate that enforces factual density and continuity, mathematically preventing the truncation or loss of critical parameters during re-indexing.
• Immunity Shield: Operational errors, tool crashes, transient hallucinations, and conversational apologies are filtered out by design and never committed to long-term memory.

Support & Feedback

Encountered an issue or have a feature suggestion?

• Bug Reports & Issues: Open an Issue
• Status & Portal: cortex.bluesynergy.io
• For a detailed theoretical and operational breakdown, see our Engineering Whitepaper (ARCHITECTURE.md).