mcpField
C
calibration-routing

calibration-routing

Remote
github.com/nao1234gAI & Agents0Updated 14w ago
Repository

About calibration-routing

calibration-routing is a read-only MCP server for checking measured judgment quality before delegating tasks. It uses third-party-resolved, sealed calibration records and lets users verify the published Brier aggregate offline.

Key features

  1. 01Query measured judgment quality
  2. 02Third-party-resolved calibration records
  3. 03Offline Brier aggregate verification
  4. 04Unverified candidates denied by default

Use cases

  1. 01Check a candidate before task delegation
  2. 02Audit published calibration records

MCP support

It exposes an HTTP route endpoint and a discovery index for MCP clients. Its calibration query supports routing tasks using measured judgment quality.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://nowpattern.com/calibration-routing/route
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

nowpattern-calibration-routing (MCP)

Query a counterparty's MEASURED judgment quality before delegating.

Third-party-resolved, OTS-sealed calibration records. Read-only. Money OFF. Default-deny on unverified candidates. The operator's own losses are published (notary honesty doctrine).

Endpoints

Zero-trust verification (3 lines, no account, offline)

curl -O https://nowpattern.com/.well-known/rjp/genuine_only_v1.json
curl -O https://nowpattern.com/.well-known/rjp/rjp_verify.py
python rjp_verify.py genuine_only_v1.json

Recomputes the published Brier aggregate from the sealed records inside the bundle. Honesty context included: genuine-only segment n=173, mean Brier 0.3096 vs 0.2117 base rate, 74.93% provenance contamination quarantined and disclosed, Hosmer-Lemeshow / Murphy-decomposition audit attached.

License

Data: CC BY 4.0 with attribution to nowpattern.com. Verifier: MIT.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.