correctover-mcp-server
Enables MCP clients to execute tool calls with CCS runtime verification across six dimensions, ensuring fail-closed conformance before any action is executed.
README
correctover-mcp-server
CCS-native MCP Server — 6-dimension runtime verification with fail-closed guarantee
What is this?
correctover-mcp-server is an MCP (Model Context Protocol) server that embeds CCS (Correctover Conformance Standard) runtime verification directly into the protocol layer. Every tool call is verified across 6 dimensions before execution — if verification fails, the action is never executed (fail-closed).
This is not an observer-pattern governance hook. It is a protocol-level guarantee.
Architecture
┌─────────────────────────────────────────────────────┐
│ MCP Client │
│ (any MCP 2026-07-28 compliant client) │
├─────────────────────────────────────────────────────┤
│ Streamable HTTP Transport │
│ (POST /mcp, _meta version negotiation) │
─────────────────────────────────────────────────────┤
│ CCS Runtime Verification │
│ ┌──────────┬──────────┬──────────┬──────────┐ │
│ │ Structure│ Schema │ Latency │ Cost │ │
│ │ Verifier │ Validator│ Monitor │ Monitor │ │
│ ├──────────┼──────────┼──────────┼──────────┤ │
│ │ Identity │ Integrity│ Policy │ Failover │ │
│ │ Tracker │ Checker │ Engine │ Engine │ │
│ └──────────┴──────────┴──────────┴──────────┘ │
│ ANY dimension fails → Action BLOCKED (never runs) │
├─────────────────────────────────────────────────────┤
│ Tool Execution │
│ (only reached if ALL 6 dimensions pass) │
└─────────────────────────────────────────────────────┘
6-Dimension Verification
| Dimension | Verifies | Failure Mode |
|---|---|---|
| Structure | Request has valid structure, protocol version, method | Malformed request rejected |
| Schema | Input matches expected tool schema | Invalid input rejected |
| Latency | Verification within budget (P50<10μs, P99<25μs) | Timeout → Fail-Closed |
| Cost | Resource usage within limits | Budget exceeded → blocked |
| Identity | Client is traceable (clientInfo present) | Untraceable → rejected |
| Integrity | Request is complete and non-empty | Corrupted → rejected |
Quick Start
Install
git clone https://github.com/Correctover/correctover-mcp-server.git
cd correctover-mcp-server
npm install
Run
# Development
npm run dev
# Production
npm run build
npm start
Configure
# Environment variables
export MCP_PORT=3000 # Default: 3000
export MCP_HOST=127.0.0.1 # Default: 127.0.0.1
export CCS_AUDIT=true # Enable audit logging
MCP Protocol Compliance
This server implements MCP 2026-07-28 (modern-only):
- ✅
server/discover— broadcasts protocol versions, capabilities, CCS identity - ✅ Streamable HTTP transport (POST /mcp)
- ✅
_metaper-request version negotiation (no initialize handshake) - ✅
resultTypeon all responses (complete/input_required) - ✅ MRTR pattern for human-in-the-loop (InputRequiredResult)
- ✅
subscriptions/listenfor change notifications - ✅
CacheableResultwithttlMs+cacheScope - ✅ Standard headers (
Mcp-Method,Mcp-Name,MCP-Protocol-Version) - ❌ No
initializehandshake (modern-only, no dual-era) - ❌ No HTTP+SSE transport (deprecated)
Available Tools
ccs_verify
Run 6-dimension runtime verification on an agent action.
{
"agent_id": "agent-123",
"action_type": "tool_call",
"tool_name": "search_web",
"tool_input": { "query": "test" }
}
ccs_evidence_hash
Generate integrity evidence hash for audit chain sealing.
{
"action_id": "action-456",
"result": { "output": "verified" },
"verifier_id": "ccs"
}
ccs_status
Get CCS runtime status and performance statistics.
Available Resources
ccs://policy/default— Default verification policy configurationccs://status/runtime— Real-time runtime statisticsccs://config/verifier— Verifier configuration (latency budget, cost limits)
CCS Extension
This server advertises the CCS extension via server/discover:
{
"extensions": {
"io.modelcontextprotocol/ccs": {
"version": "1.0.0",
"dimensions": ["structure", "schema", "latency", "cost", "identity", "integrity"],
"failClosed": true,
"maxOverheadUs": 25
}
}
}
Integration with Halo
The ccs_evidence_hash tool generates hashes compatible with Halo's verification block (v0.2.30+), enabling sealed audit chains:
{
"verification": {
"status": "verified",
"verifier": "ccs",
"policy_ref": "default",
"checked_at": "2026-08-03T12:00:00Z",
"evidence_hash": "0x..."
}
}
Performance
- P50 verification latency: < 10μs
- P99 verification latency: < 25μs
- Fail-closed guarantee: Action blocked if ANY dimension fails
Development
# Build
npm run build
# Test
npm test
# Lint
npm run lint
# Format
npm run format
License
MIT
References
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
E2B
Using MCP to run code via e2b.