MCP 服务器

FactorGuide

Accepts a coupling matrix (covariance, correlation, or precision) and returns zone classifications, optimal factorization strategy, and calibrated risk predictions. Answers: which variable dependencies are load-bearing and which can be safely ignored?

README

FactorGuide

Decision intelligence for AI agents. Send a coupling matrix — get zone classifications, optimal factorization strategy, and calibrated risk predictions.

Every system with interacting variables has a coupling structure. When an agent simplifies that system by treating variables as independent, it pays an information cost. FactorGuide quantifies that cost exactly.

Quick Start

MCP (Model Context Protocol)

FactorGuide is listed on the Official MCP Registry. Any MCP-compatible agent discovers tools automatically:

mcp_endpoint: https://factorguide.io/mcp

HTTP REST

POST https://factorguide.io/navigate
POST https://factorguide.io/diagnose
POST https://factorguide.io/explain
POST https://factorguide.io/report_outcome

Agent Discovery

GET https://factorguide.io/llms.txt
GET https://factorguide.io/openapi.json

Example

Send a 3×3 correlation matrix:

{
  "coupling": {
    "covariance_matrix": [
      [1.00, 0.72, 0.05],
      [0.72, 1.00, 0.48],
      [0.05, 0.48, 1.00]
    ]
  },
  "sample_size": 500,
  "model_class": "constitutive"
}

Get back zone classifications:

Pair	IC	Zone	Recommendation
z0–z1	0.682	3	PRESERVE
z0–z2	0.031	1	FACTORIZE
z1–z2	0.453	2	ASSESS

Zone 1: Safe to factorize. Coupling below threshold.
Zone 2: Depends on functional role. Same IC, opposite recommendations for constitutive vs. inductive coupling.
Zone 3: Must preserve. Coupling is load-bearing regardless of model class.

Pricing

Tier	Price	Queries	Max Variables
Trial	Free	15 per wallet	n ≤ 25
Starter	$0.05/query	50–200 bundles	n ≤ 100
Professional	$0.03/query	500–2000 bundles	n ≤ 1000

Payment: USDC on Base via x402 or MPP. report_outcome is always free.

Data Privacy

FactorGuide accepts only second-order summary statistics — covariance, precision, correlation matrices, or edge lists. No raw observations. Matrices are zeroed after IC computation.

Theoretical Foundation

Built on Circulatory Fidelity, a mathematical framework where IC (Inference Coupling) measures the partial correlation between variables. The cost function V(IC) gives the exact mutual information destroyed by severing a coupling. Risk curves are calibrated on 49,000+ validated datapoints.

Every prediction is falsifiable. When agents report outcomes, the flywheel refines future predictions.