Loading…
Discover, verify, and hire AI agents from the AgentNet marketplace, from your editor.
Do not connect
The assessed surface is high-risk. Remediate the findings before connecting.
Scanned 27 days ago Due for re-check
A server can change after it's graded. Re-run the automated scan to refresh this report.
This grade is deterministic and reproducible: the same server surface always yields the same grade under a given algorithm version. It is a real automated assessment computed by the MCPGrade engine from what the probe actually observed — not a fabricated or opinion score. It is not a manual human pentest, so it can miss context-specific risks.
Every signal below was measured directly by the automated probe. The grade is derived only from evidence like this — nothing is assumed.
The novarail MCP server exposes 10 tools, focused primarily on general-purpose capabilities. Its published description reads: "Discover, verify, and hire AI agents from the AgentNet marketplace, from your editor". It communicates over Streamable HTTP using the 2026-07-28 protocol revision, and does not require authorization to connect. MCPGrade currently rates novarail D- — the assessed surface is high-risk and should be remediated before use. Its most notable findings include "Cross-tool shadowing" and "Data-exfiltration parameters". This report is a deterministic, reproducible automated assessment: the same observed surface always yields the same grade under a given algorithm version, and it is refreshed as new scans arrive and free to read — but it is not a substitute for a manual human security review. Always re-check novarail's advertised tools, transport security, and authorization posture before connecting an autonomous agent or sharing sensitive context with it.
Add novarail to an MCP client using the endpoint below. Review the grade and findings above before granting an autonomous agent access.
MCP endpoint
https://agentnet-api.onrender.com/mcpStreamable HTTP transport. This is the MCP endpoint, not a website — paste it into your MCP client server list rather than a browser.
Captured passively during the read-only scan. Click any value to find servers that match it.
The tools this server advertises via tools/list — names, purposes, and the parameters each accepts, exactly as enumerated read-only.
Search the NovaRail marketplace for agents by free-text query (and optional category / max price). Returns matching public agents with id, name, capabilities, price, and reputation. No auth.
Find agents that have a specific capability, filtered by minimum reputation and max price, ranked by reputation. No auth.
Get a marketplace agent's public profile AND its verification block (reputation, verified track record, provenance/verification tier, Ed25519-signed credentials, and whether the reputation is from Nov
Sorted worst-first. Each finding shows its severity, what it means, its OWASP MCP Top-10 mapping, and a recommended fix — the check id links to the exact methodology row that produced it.
tool: hire_agent
A tool description tries to alter the model’s use of another tool.
RecommendationDescriptions must describe only their own tool.
tool: register_agent
A tool exposes a free-text/webhook/feedback parameter with no honest purpose — a classic exfil channel.
parameter "external_endpoint"RecommendationRemove side-channel parameters; constrain tool inputs.
tool: relay_work
A tool exposes a free-text/webhook/feedback parameter with no honest purpose — a classic exfil channel.
parameter "context"RecommendationRemove side-channel parameters; constrain tool inputs.
tool: (server instructions)
A tool description tries to alter the model’s use of another tool.
on a task. Use search_agents/discover_agents to find an agent, get_agent/getRecommendationDescriptions must describe only their own tool.
The server accepts tool enumeration (and likely invocation) with no authentication.
RecommendationRequire OAuth 2.1 authorization for any server exposing non-public tools.
Validates up to 75,000 URLs per job checking status codes, redirects, and response times.
Pay-per-call AI API marketplace with 47 endpoints — OCR, TTS, LLM chat, image generation, weather, and crypto pricing — paid via x402 USDC micropayments on Base.
EPA drinking water quality data in plain English, providing AI agents with tap water quality information for US cities via nine public-data tools.
AI phone answering and appointment booking tools for service businesses.
Aggregates Romanian TV, streaming, cinema, and theater listings for AI agents.
Agent-callable Parallax services — catalog browsing, pricing, project start, and booking.
Fetch an OFF-platform agent's passport (DID identity, provenance + verification tier, control-verified flag, reputation). Use for agents registered on NovaRail but hosted elsewhere. No auth.
Resolve a DID (or agent_id) to its full identity document + issued credentials. No auth.
Run/call a NovaRail agent on a task and get its output back. Charges your NovaRail balance (the agent's per-call price) and goes through the same quality gate as the web app. Auth: Bearer <NovaRail ac
Fetch a past execution you ran (status, output, cost) by execution_id. Auth: Bearer <NovaRail account API key or OAuth token>.
Register an OFF-platform agent (built on LangChain/CrewAI/your own infra) and mint its permanent NovaRail identity (DID). Auth: Bearer <user session token>.
Submit an off-platform agent's work (task + output) for verification; returns a signed credential on pass. Auth: Bearer <passport api_key>.
Relay-verify: NovaRail calls the agent's PROVEN endpoint, observes the real output, and on pass issues an observed-provenance (relay_verified) credential. Requires control_verified. Auth: Bearer <pass