Developer Infrastructure & Products

Developer Documentation & Technical Reference

Complete technical specifications, quickstart installation guides, and configuration references for integrating Neon's Model Context Protocol (MCP) servers and security gateways into your development environment.

Executive Architecture Summary (AEO Definition Block)
Anthropic MCP • JSON-RPC 2.0 • Linux Foundation A2A Standard

Neon Innovation Lab Developer Documentation provides normative engineering references for deploying our 22+ Model Context Protocol (MCP) servers, zero-trust DLP proxies, and agentic infrastructure. Designed for software engineers, security architects, and CTOs, this documentation covers Stdio and SSE transports, configuration schemas, and SARIF 2.1.0 security compliance.

JSON-RPC 2.0
Transport Spec
Anthropic protocol standard
< 5 Min
Quickstart Setup
1-click copyable config
Stdio & SSE
Execution Modes
Local CLI and Cloud endpoints
SARIF 2.1
Security Output
GitHub Security Gate ready
Core Engineering Capabilities

Architectural Specifications & Systems

Deterministic execution patterns, verified protocol compliance, and security controls built for production environments.

01

Claude Desktop Configuration Standard

Adding Neon zero-trust gateways and security auditors to your local Claude Desktop client via stdio.

{ "mcpServers": { "neon-zero-trust-gateway": { "command": "npx", "args": ["-y", "@neon-innovation-lab/zero-trust-gateway"], "env": { "GATEWAY_ENFORCE_GDPR": "true", "STRIP_API_KEYS": "true" } } } }
  • Config location (macOS): ~/Library/Application Support/Claude/claude_desktop_config.json
  • Config location (Windows): %APPDATA%\Claude\claude_desktop_config.json
  • Automatic environment variable validation and failover handling
02

Cursor IDE SSE Gateway Integration

Connecting Cursor IDE to cloud-multiplexed MCP servers to eliminate local context bloat and bypass tool count caps.

{ "mcp": { "servers": [ { "name": "neon-universal-hub", "type": "sse", "url": "https://neoninnovationlab.com/api/v1/mcp" } ] } }
  • Config location: ~/.cursor/mcp.json
  • Single multiplexed endpoint streaming tool definitions via Server-Sent Events
  • Dynamic tool activation preserving IDE prompt token budgets
03

A2A Agent Card Security Scanning in CI/CD

Automating Agent2Agent manifest verification within GitHub Actions to prevent prompt injection and unauthorized capabilities.

# .github/workflows/a2a-audit.yml - name: Audit Agent Card run: npx -y a2a-audit scan ./.well-known/agent-card.json --sarif results.sarif
  • Offline Ed25519 digital signature validation
  • SSRF defense against private network and cloud metadata leaks
  • SARIF 2.1.0 output automatically displayed in GitHub Security tab
04

Environment Variables & DLP Secret Filters

Fine-tuning in-band DLP inspection policies to match your organization's specific data classification and privacy standards.

  • BLOCK_DESTRUCTIVE_SQL: Intercepts unindexed DELETE, DROP, and TRUNCATE
  • STRIP_API_KEYS: Redacts AWS, OpenAI, Stripe, and GitHub secret tokens
  • GATEWAY_ENFORCE_GDPR: Masks emails, phone numbers, and government IDs
Performance & Rigor Audit

Integration Transport Comparison

Direct side-by-side evaluation against typical industry implementations.

FeatureStdio Transport (Local npx)SSE Transport (Managed Cloud)
Execution EnvironmentRuns locally on developer machine via Node.jsRuns on high-availability cloud serverless clusters
Network OverheadZero network latency; runs via standard pipesSub-50ms P95 latency over encrypted HTTPS/WSS
Multi-Tool MultiplexingEach server runs as an independent child processUnified single endpoint multiplexing 50+ tools
Context Token UsageLoads full schema for each active serverDynamic tool routing saving up to 89% of tokens
Structured Engagement Roadmap

Four-Stage Engineering Pipeline

Every technical engagement follows rigorous architecture, sandboxed validation, and production rollout criteria.

01

Select Transport Mode

Choose between local stdio execution for offline privacy or managed SSE for team multiplexing.

Milestone: Target Integration Configuration File
02

Define DLP Policies

Configure environment variables for secret redaction, SQL circuit breakers, and privacy enforcement.

Milestone: Hardened Security Configuration Profile
03

Test Tool Invocation

Trigger test commands in Claude or Cursor to verify tool parameter schemas and response handling.

Milestone: Verified Interactive Tool Execution
04

Automate CI/CD Auditing

Incorporate static AST scanners and A2A card verifiers into pull request verification workflows.

Milestone: Continuous Integration Security Gate
Technical Knowledge Base

Frequently Addressed Questions

Concrete answers covering architecture, integration protocols, security, and SLAs.

Where do I report documentation issues or suggest improvements?+

All developer documentation is maintained in our GitHub repository. Pull requests and issues are reviewed directly by our engineering principals.

What version of Node.js is required for local MCP servers?+

All Neon MCP tools require Node.js v18.0.0 or higher. Node.js v22 LTS is recommended for optimal performance.

Does the Zero-Trust Gateway log prompt contents?+

No. The gateway processes data entirely in-band and in-memory. Zero prompt text or tool outputs are logged to disk or sent to external servers in the open-source tier.

Enterprise Deployment • Dubai Regional HQ • Global Operations

Ready to Deploy Production Architecture?

Connect with our principal engineering team in Dubai to discuss technical requirements, security boundaries, and delivery schedules.