Injection Detector is a remote MCP server that gives an AI agent 2 tools, among them detect_injection, detect_trace_tool_policy. It answers at https://mcp.viridis-security.com/mcp over streamable HTTP, and it listed its tools to an anonymous client, so no sign-in was needed for that. Our probes reached it 6 of 6 times since 11 September 2026; the last check, on 24 October 2026, succeeded.
Formally-verified injection/exfiltration detector for AI agents (MCP-02).
Description as the publisher wrote it.
| Tool | What it does |
|---|---|
detect_injection | Screen untrusted input for prompt/tool injection, exfiltration, and obfuscation before an agent consumes it. Returns a verdict (clean|suspicious|attack), probability, bits-at-risk (upper bound on adversarial capture per the Adversarial Land Inputs: input, context, certainty, agentId |
detect_trace_tool_policy | Analyze an agent trace for the Gray Swan Wave 16 class: untrusted retrieved/tool output causing a tool call outside the user-declared per-turn allowlist. Returns trace counts, unauthorized tool-call evidence, canon mapping VC-AI-TOOL-0001, Inputs: trace, traces, targetName |
Every block below points at https://mcp.viridis-security.com/mcp. Pick your client and copy.
claude mcp add --transport http injection-detector https://mcp.viridis-security.com/mcp
{
"mcpServers": {
"injection-detector": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://mcp.viridis-security.com/mcp"]
}
}
}{
"mcpServers": {
"injection-detector": {
"url": "https://mcp.viridis-security.com/mcp"
}
}
}{
"servers": {
"injection-detector": {
"type": "http",
"url": "https://mcp.viridis-security.com/mcp"
}
}
}{
"context_servers": {
"injection-detector": {
"url": "https://mcp.viridis-security.com/mcp"
}
}
}{
"mcpServers": {
"injection-detector": {
"serverUrl": "https://mcp.viridis-security.com/mcp"
}
}
}{
"mcpServers": {
"injection-detector": {
"url": "https://mcp.viridis-security.com/mcp"
}
}
}name: Injection Detector
version: 0.0.1
schema: v1
mcpServers:
- name: Injection Detector
type: streamable-http
url: https://mcp.viridis-security.com/mcp{
"mcpServers": {
"injection-detector": {
"httpUrl": "https://mcp.viridis-security.com/mcp"
}
}
}{
"mcpServers": {
"injection-detector": {
"type": "http",
"url": "https://mcp.viridis-security.com/mcp"
}
}
}{
"mcpServers": {
"injection-detector": {
"url": "https://mcp.viridis-security.com/mcp"
}
}
}goose://extension?url=https%3A%2F%2Fmcp.viridis-security.com%2Fmcp&type=streamable_http&id=injection-detector&name=Injection%20Detector
It answered our last check on 24 October 2026 in 432 ms. Over 6 checks since 11 September 2026 it answered 6. Status here comes from Neuronto's own probes, not from the publisher.
2, as returned by the server's own tools/list on 22 September 2026: detect_injection, detect_trace_tool_policy.
It answered our last check on 24 October 2026 in 432 ms. Over 6 checks since 11 September 2026 it answered 6.
Put the server under mcpServers in ~/.cursor/mcp.json with the URL https://mcp.viridis-security.com/mcp. The exact JSON is above, with the same setup for Claude Code, Claude Desktop, VS Code, Windsurf, Zed, Gemini CLI and the rest.
It listed its tools to an anonymous client, so reading the tool list needs no key. Individual tools can still ask for credentials in their own inputs.
Publisher: io.github.viridis-security. Endpoint host: mcp.viridis-security.com. ARD identifier: urn:air:io.github.viridis-security:mcp:injection-detector.
First indexed 10 September 2026. The tools are exactly what the server returned; nothing on this page
is inferred from its name. Ranking in Neuronto search is never for sale.