sebbi.pro is an MCP server — so anything that speaks MCP connects to it with a single link. Seal decisions, check sources, rate insurability and run all 755 compliance checks from inside whatever you already build with.
https://sebbi.pro/mcpPaste this wherever a tool asks for an MCP server. That's the whole setup.
One command in your terminal:
claude mcp add --transport http sebbi https://sebbi.pro/mcpIn your workspace: Add connector → Custom MCP server, paste the link. Now every Lovable build can seal and check against sebbi.
https://sebbi.pro/mcpIn an MCP-enabled Space or agent, add an MCP server with the link. Your Space can then call sebbi's tools directly.
https://sebbi.pro/mcpAny MCP-capable editor or agent: add an HTTP MCP server with the link. Same address, same tools.
https://sebbi.pro/mcpWhere connectors / MCP are available, add the link as a custom connector. For apps that call the API directly, use the gateway below.
https://sebbi.pro/mcpIf your app calls a model API instead of using MCP, route it through sebbi with a one-line base-URL change and every call is governed and sealed. Full setup on /connect and /build. Every governed response carries an AI-Decision-Receipt header — now submitted to the IETF as an Internet-Draft.
Four calls any agent can make, from any framework, whether or not it speaks MCP:
Stop button — check before every action: GET https://sebbi.pro/x/jugular/check?id=JG-… (/jugular)
Plan Lock — lock the plan, then check each tool call: POST /x/planlock/commit, /x/planlock/check (/plan-lock)
Mandate Chain — authority that only narrows: POST /x/mandate/issue, /delegate, /check (/mandate)
Oversight Keys — wait for a person: POST /x/oversight_keys/open (/oversight-keys)
Same address, same tools and all 755 checks, everywhere. New to it? Start at /connect · see what it guards at /checks · the standard at /insurable.