iTechSmart MCP Server
Official> [!WARNING]
> **This repository has moved to [github.com/iTechSmartINC/mcp-server](https://github.com/iTechSmartINC/mcp-server)** as part of the iTechSmart Inc rebrand.
> Please update your links and git remotes. This copy is kept as a redirect and is no longer updated.
# iTechSmart MCP Server
**Connect any AI agent to UAIO — Autonomous IT Operations + ProofLink™ Verification**
> The SaaS products that become agent-ready in 2026 win the enterprise deals.
---
## What this does
This MCP server exposes iTechSmart's UAIO platform as tools any MCP-compatible AI agent can call — Claude, ChatGPT, Copilot, Cursor, and any other agent that speaks the Model Context Protocol.
Once connected, your AI agent can:
- **Verify** ProofLink cryptographic receipts — confirm autonomous AI actions weren't tampered with
- **Query** UAIO platform status — real-time health across 131 production containers
- **Fetch** incident details — what the AI did, when, with cryptographic proof
- **List** recent autonomous remediations — full audit trail
- **Simulate** infrastructure attacks — trigger the break-it sandbox live
---
## Quick Start
### Claude Desktop
Add to `~/Library/Application Support/Claude/claude_desktop_config.json`:
```json
{
"mcpServers": {
"itechsmart": {
"command": "npx",
"args": ["-y", "@itechsmart/mcp-server"],
"env": {
"ITECHSMART_API_KEY": "your-api-key-here"
}
}
}
}
```
Get your API key: contact [enterprise@itechsmart.dev](mailto:enterprise@itechsmart.dev) or [itechsmart.dev/contact](https://itechsmart.dev/contact)
### Remote MCP (SSE)
```
https://mcp.itechsmart.dev/sse
```
For platforms that support remote MCP servers directly (Cursor, Continue, etc.).
---
## Tools
### `verify_prooflink_receipt`
Verify the cryptographic integrity of a ProofLink receipt.
```
Agent: "Verify ProofLink receipt a1b2c3d4e5f6a7b8"
→ Returns:
{
"receipt_id": "a1b2c3d4e5f6a7b8",
"verification_result": "VERIFIED ✓",
"tamper_detected": false,
"action": "kubectl patch memory 512Mi→1024Mi + rollout restart",
"human_input": "ZERO",
"sha256": "abc123...",
"verify_url": "https://verify.itechsmart.dev/a1b2c3d4e5f6a7b8"
}
```
### `get_receipt_chain`
Fetch and verify the complete ProofLink receipt chain.
```
Agent: "Check the full UAIO audit trail for the last 20 actions"
→ Returns: complete chain with tamper detection across all receipts
```
### `query_uaio_status`
Get real-time UAIO platform health and metrics.
```
Agent: "What's the current status of the iTechSmart platform?"
→ Returns:
{
"status": "OPERATIONAL",
"containers": { "healthy": 131, "total": 131 },
"prooflink": { "receipts_generated": 134, "chain_breaks": 0 },
"compliance": { "nist_csf": "96/100", "hipaa": "100/100" }
}
```
### `get_incident_details`
Fetch full details of a specific autonomous remediation.
```
Agent: "Show me what UAIO did during incident inc-1746960000"
→ Returns: trigger, action taken, before/after state, ProofLink receipt
```
### `list_recent_incidents`
List recent autonomous IT remediations.
```
Agent: "List the last 10 things UAIO fixed automatically"
→ Returns: chronological incident list with detection/remediation times
```
### `simulate_infrastructure_attack`
Trigger the break-it sandbox — watch the UAIO loop live.
```
Agent: "Simulate a Kubernetes OOMKilled crash and show me the ProofLink receipt"
→ Returns: full 5-phase UAIO simulation with cryptographic receipt
```
---
## Example Agent Workflows
### Compliance Audit Workflow
```
"Fetch the last 50 ProofLink receipts, verify chain integrity,
and generate a compliance report for our SOC 2 audit."
→ Agent calls get_receipt_chain(limit=50)
→ Verifies all 50 receipts
→ Reports: chain_valid, any tamper positions, NIST control coverage
```
### Incident Investigation Workflow
```
"What did the AI fix last night between 2-4AM and can you prove it?"
→ Agent calls list_recent_incidents(since="2026-05-11T02:00:00Z")
→ Agent calls get_incident_details for each incident
→ Agent calls verify_prooflink_receipt for each receipt
→ Returns: complete verified audit trail
```
### Monitoring Workflow
```
"Check if the iTechSmart platform is healthy before we run our demo."
→ Agent calls query_uaio_status()
→ Returns: operational status, container health, compliance scores
```
---
## Environment Variables
| Variable | Required | Description |
|----------|----------|-------------|
| `ITECHSMART_API_KEY` | Yes | API key — contact enterprise@itechsmart.dev or itechsmart.dev/contact |
| `ITECHSMART_API_URL` | No | Custom API URL (default: https://app.itechsmart.dev/api/v1) |
---
## Development
```bash
git clone https://github.com/Iteksmart/mcp-server
cd mcp-server
npm install
npm run build
npm start
```
---
## Links
- Platform: [itechsmart.dev](https://itechsmart.dev)
- Live demo: [itechsmart.dev/break-it](https://itechsmart.dev/break-it)
- Verify receipts: [verify.itechsmart.dev](https://verify.itechsmart.dev)
- Open-source verifier: [github.com/Iteksmart/prooflink-verifier](https://github.com/Iteksmart/prooflink-verifier)
- API docs: [docs.itechsmart.dev](https://docs.itechsmart.dev)
---
## About iTechSmart
iTechSmart builds UAIO — Unified Autonomous IT Operations. The first platform that autonomously detects, remediates, and cryptographically proves every infrastructure action.
SDVOSB · CAGE: 172W2 · NVIDIA Inception · F6S #6 Global · NIST CSF 96/100
---
MIT License — iTechSmart Inc. 2026
TDQS
Scored across 66 tools
Many tools overlap in purpose, e.g., multiple incident listing tools (list_recent_incidents, mission.list_incidents, itsm_list_open_incidents) and numerous health check tools (langfuse_health, mesh_health, probo_health, etc.). Aliases like compliance.audit_summary for get_compliance_status add confusion.
Tools use both snake_case (get_compliance_status) and dot notation (mission.list_incidents, prooflink.verify_receipt). Different prefixes (itsm_, ghl_, trmm_) are consistent within their subgroup but overall pattern is mixed.
66 tools is excessive for a single server. The server attempts to cover multiple unrelated domains (ITSM, GHL, Deepgram, compliance, integrations), lacking a focused scope. Many tools are redundant or too granular.
The tool set covers many read operations but lacks write operations for several domains (e.g., no update_ticket in ITSM). There are many health checks but limited action tools. The domain is poorly defined, making it hard to assess completeness.