This week:
3660 new agents scanned
Grade Distribution
Full reportRecently Scanned
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Ronnietag/Linux MCP Server
tgosoul2019/Multi-LLM Gateway MCP
scamverifyai/ScamVerify
Vicus-Brits/MCP Database Query Server
Splaticusamz/Shopify MCP Server
nickedridge-wq/Roboflow MCP Server
cyc20020404-sys/Weather MCP Server
0xHashy/fathom-fyi
Needs Improvement
View all
D
D
D
D
D
D
D
D
com.aineedhelpfromotherai/reasoning-commons
dev.whatcanido.www/mcp
io.github.b-shields/cmxflow
io.github.bch1212/injectshield
ai.compeller/compel
io.github.myrouteonline/myrouteonline
io.github.atmflow55/datafood-mcp
io.github.juanisidoro/securecode
How AgentGrade Works
AgentGrade scans MCP servers and AI agent endpoints with passive HTTP checks — no authentication attempts, no exploitation. Every agent gets a letter grade from A to F based on 14 security criteria including HTTPS, authentication, CORS configuration, credential exposure, and security headers.
Why MCP Server Security Matters
The Model Context Protocol (MCP) connects AI assistants to external tools and data. A misconfigured MCP server can expose credentials, leak data through CORS misconfigurations, or allow unauthorized access. AgentGrade helps developers and teams verify their agent’s security posture before deployment.