EVIDIQ Sentinel

Introspects a public remote MCP endpoint and its payment discovery without invoking any target tool, then returns a signed security report scored from 0 to 100. Provide: 1. the endpoint URL 2. optionally whether to include x402 payment discovery.

Endpoint:https://mcp.evidiq.dev/sentinel/mcp

Audits a supplied MCP server manifest for dangerous tool exposure, prompt-injection phrases, and excessive permissions. Provide: 1. the manifest JSON.

Endpoint:https://mcp.evidiq.dev/sentinel/mcp

Audits an Agent Skill document for injection, credential exfiltration, and unsafe instruction patterns before an agent loads it. Provide: 1. the Skill markdown or its URL.

Endpoint:https://mcp.evidiq.dev/sentinel/mcp

Scans a combined bundle of endpoint, manifest, and Skill inputs in one pass and returns a single consolidated verdict. Provide: 1. the bundle object.

Endpoint:https://mcp.evidiq.dev/sentinel/mcp

Returns supported scan targets, safety limits, detection rule families, and full pricing. Provide: no input is required.

Endpoint:https://mcp.evidiq.dev/sentinel/mcp

Validates a scan target and its DNS safety before payment, without contacting the target. Provide: 1. the target 2. the target type.

Endpoint:https://mcp.evidiq.dev/sentinel/mcp

Returns the exact fixed price for a selected paid tool. Provide: 1. the name of the paid tool.

Endpoint:https://mcp.evidiq.dev/sentinel/mcp

Recomputes a report hash and verifies its EIP-191 signature against the expected signer. Provide: 1. the report 2. its signature 3. the signer address.

Endpoint:https://mcp.evidiq.dev/sentinel/mcp