Agentfox
MCP server · architsharm
Connect
Claude Code
claude mcp add architsharm-agentfox -- uvx agentfox
Config file (Cursor, VS Code, Claude Desktop …)
{
"mcpServers": {
"architsharm-agentfox": {
"command": "uvx",
"args": [
"agentfox"
]
}
}
}
MCP servers can read and change data in the connected service. Only connect servers you trust and check which permissions you grant.
What does the Agentfox MCP server do?
Questions about Agentfox
What does the Agentfox MCP server do?
Open source control plane for AI agents: least privilege on tool calls and a checkable audit trail.
How do I connect Agentfox?
It runs locally. In Claude Code: claude mcp add architsharm-agentfox -- uvx agentfox. Other clients take the same command in their MCP config file.
Is Agentfox an official MCP server?
Not confirmed: it is published by architsharm, not necessarily by the company behind the connected service. Check the source before you connect it.
More MCP servers: Security
Malware Patrol is a remote MCP server (Threat Intelligence) run by Malware Patrol. Add it as a connector to let Claude, ChatGPT or Cursor work with Malware Patrol directly. Sign-in: API Key.
List devices, alerts, antivirus status and audit logs, and run scripts on devices in Pulseway.
Can ChatGPT, Claude and Perplexity crawlers read your page? Audit, fixes, compare, monitoring.
Managed Keycloak from any MCP client: clusters, realms, apps, SSO, users, domains, audit events.
Audit ERC-20 honeypot & buy/sell tax risks before routing non-custodial 0.25% EVM swaps.
Live GEO readiness audit: check JSON-LD schemas, page speed, and AI crawler robots.txt access.
Threat intelligence: threats, detections (SPL/KQL/Sigma), IOCs, CVEs/CWEs, ATT&CK coverage, actors
Live checks: AI crawler access, llms.txt, entity schema and AI-search readiness of any site.