How do I use graph-attention-engine?

This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.

graph-attention-engine

by ArindamBanerji

Overview Schema Source Related Servers Discussions Admin

Hybrid

Graph Attention Engine (GAE)

Graph Attention Engine is a standalone, NumPy-only scoring engine that implements Cross-Graph Attention for domain-independent decision systems. It is pure Python and requires no database, network access, or event loop. GAE is licensed under Apache 2.0.

Install

Install the published package:

pip install graph-attention-engine

For editable development installs:

pip install -e ".[dev]"

Related MCP server: coord-mem MCP Server

Quick Start

Scaled dot-product attention

import numpy as np
from gae import scaled_dot_product_attention

rng = np.random.default_rng(7)
Q = rng.normal(size=(4, 8))
K = rng.normal(size=(6, 8))
V = rng.normal(size=(6, 16))

out, weights = scaled_dot_product_attention(Q, K, V)

assert out.shape == (4, 16)
assert weights.shape == (4, 6)

Profile scoring

ProfileScorer compares a factor vector with category-specific action centroids. The convenience factory accepts named categories and actions:

import numpy as np
from gae import ProfileScorer, build_profile_scorer

categories = ["routine", "critical"]
actions = ["accept", "review", "escalate"]

# Shape: (2 categories, 3 actions, 4 factors)
mu = np.array(
    [
        [
            [0.2, 0.2, 0.1, 0.2],
            [0.5, 0.4, 0.5, 0.4],
            [0.8, 0.8, 0.9, 0.8],
        ],
        [
            [0.1, 0.2, 0.2, 0.1],
            [0.5, 0.6, 0.5, 0.6],
            [0.9, 0.9, 0.8, 0.9],
        ],
    ],
    dtype=np.float64,
)

centroids = {
    category: {
        action: mu[category_index, action_index].tolist()
        for action_index, action in enumerate(actions)
    }
    for category_index, category in enumerate(categories)
}

scorer: ProfileScorer = build_profile_scorer(
    categories=categories,
    actions=actions,
    centroids=centroids,
    n_factors=4,
)
result = scorer.score(np.array([0.85, 0.90, 0.80, 0.90]), category_index=1)

print(result.action_index)
print(result.confidence)
print(result.probabilities)

The returned ScoringResult contains the recommended action_index, action_name, calibrated confidence, full probabilities, distances, entropy, and confidence gap.

Core API

Core Scoring

API	Description
`ProfileScorer`	Scores factor vectors against category and action centroids.
`ScoringResult`	Contains the recommendation, confidence, probabilities, and diagnostics.
`KernelType`	Selects the distance kernel used for scoring.
`build_profile_scorer`	Builds a scorer from named categories, actions, and centroid mappings.

Oracle

API	Description
`OracleProvider`	Protocol for producing independent ground-truth outcomes.
`GTAlignedOracle`	Returns outcomes aligned with supplied ground truth.
`BernoulliOracle`	Samples deterministic-seeded Bernoulli outcomes.

Evaluation

API	Description
`EvaluationScenario`	Defines one scored evaluation case and expected action.
`compute_ece`	Computes expected calibration error.
`run_evaluation`	Evaluates a scorer across a scenario collection.

Judgment

API	Description
`compute_judgment`	Converts a scoring result into an explained judgment.
`CONFIDENCE_HIGH`	Threshold for high-confidence judgments.
`CONFIDENCE_MEDIUM`	Threshold for medium-confidence judgments.

Calibration

API	Description
`CalibrationProfile`	Holds scoring and learning calibration parameters.
`compute_theta_min`	Computes the deployment-aware conservation threshold.

Primitives

API	Description
`scaled_dot_product_attention`	Computes scaled dot-product attention and its weights.
`softmax`	Computes a numerically stable probability distribution.

Learning

API	Description
`LearningState`	Tracks centroids and learning metadata.
`WeightUpdate`	Describes a validated learning update.

Convergence

API	Description
`get_convergence_metrics`	Reports centroid movement and convergence diagnostics.

Kernels

API	Description
`L2Kernel`	Computes standard squared Euclidean distance.
`DiagonalKernel`	Applies per-factor diagonal weights.
`CovarianceEstimator`	Estimates covariance for adaptive distance calculations.
`KernelSelector`	Recommends a kernel from observed data properties.

Persistence

API	Description
`save_state`	Saves learning state to JSON.
`load_state`	Loads learning state from JSON.

The complete public export list is available in gae/__init__.py.

Conservation Law

GAE checks the constraint alpha * q * V >= theta_min, combining category coverage (alpha), observed quality (q), and verified volume (V). Meeting the threshold provides the operational guarantee that evidence is sufficient to maintain accuracy above the configured boundary; use compute_theta_min to derive the deployment-aware threshold.

MCP Server

Graph Attention Engine includes a built-in Model Context Protocol (MCP) server that exposes decision scoring as MCP tools. Any MCP-compatible client (Claude Desktop, Cursor, etc.) can connect and use GAE's cross-graph attention scoring directly.

Install

pip install "graph-attention-engine[mcp]"

Usage

# stdio transport (for Claude Desktop, Cursor)
rgi-mcp-server --categories alert_triage,incident_response --actions investigate,escalate --factors severity,confidence,blast_radius

# HTTP transport (for remote/hosted deployments)
rgi-mcp-server --http --port 8080 --categories alert_triage,incident_response --actions investigate,escalate --factors severity,confidence,blast_radius

Docker

docker build -f Dockerfile.mcp -t gae-mcp .
docker run -p 8080:8080 gae-mcp

MCP Tools

Tool	Description
`score_decision`	Score a decision using cross-graph attention and compounding intelligence
`get_trajectory`	Retrieve the learning trajectory showing how decisions compound over time
`get_conservation_status`	Check conservation law compliance and scoring integrity

Claude Desktop Configuration

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "graph-attention-engine": {
      "command": "rgi-mcp-server",
      "args": ["--categories", "alert_triage,incident_response", "--actions", "investigate,escalate", "--factors", "severity,confidence,blast_radius"]
    }
  }
}

MCP Server

Install

pip install "graph-attention-engine[mcp]"

Usage

# stdio transport (for Claude Desktop, Cursor)
rgi-mcp-server --categories alert_triage,incident_response --actions investigate,escalate --factors severity,confidence,blast_radius

# HTTP transport (for remote/hosted deployments)
rgi-mcp-server --http --port 8080 --categories alert_triage,incident_response --actions investigate,escalate --factors severity,confidence,blast_radius

Docker

docker build -f Dockerfile.mcp -t gae-mcp .
docker run -p 8080:8080 gae-mcp

MCP Tools

Tool	Description
`score_decision`	Score a decision using cross-graph attention and compounding intelligence
`get_trajectory`	Retrieve the learning trajectory showing how decisions compound over time
`get_conservation_status`	Check conservation law compliance and scoring integrity

Claude Desktop Configuration

Add to your claude_desktop_config.json:

{
  "mcpServers": {
    "graph-attention-engine": {
      "command": "rgi-mcp-server",
      "args": ["--categories", "alert_triage,incident_response", "--actions", "investigate,escalate", "--factors", "severity,confidence,blast_radius"]
    }
  }
}

Testing

pytest tests/ -v

The current suite contains 1,244 passing tests.

Documentation

License

Apache 2.0. See the LICENSE file.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

This server cannot be deployed

Claude Code, inside the Godot 4 editorFlockbay puts an AI agent in the real Godot 4 editor: it writes GDScript, runs your game and fixes what breaks. Free, or bring your own Claude Code or Codex.flockbay.comAd

Maintenance

ActivityMaintained

ResponsivenessOutdated data

Resources

Related MCP Connectors

Reasoning Commons
io.github.chenyuan35
AI Reasoning Cache & Consensus Layer with 11 MCP tools via Streamable HTTP.
AI & Machine Learning Knowledge & Memory Autonomous Agents
2
https://aineedhelpfromotherai.com/api/mcp
strata
ai.klavis
MCP server for progressive tool usage at any scale (see https://klavis.ai)
Agent Orchestration Autonomous Agents Developer Tools
25,748
https://strata.klavis.ai/mcp/
contextq-mcp
io.github.contextq
Persistent memory, hybrid search and a goal graph for AI agents, over stdio or remote HTTP.
Knowledge & Memory Search RAG Systems
1
https://app.contextq.dev/mcp
Meeba BrainA
io.github.meemoprasad
Repository knowledge graph MCP server for codebase understanding and debugging.
Code Analysis Developer Tools
51
https://meeba-context-mcp-469248886325.us-central1.run.app/mcp

Related MCP Servers

intelligraph-mini
Code Analysis Developer Tools Knowledge & Memory
kfireew
F
license
A
quality
D
maintenance
Local-first MCP graph intelligence server providing RRF hybrid search, multi-hop traversal, source snippets, and rationale nodes for AI agents, without Docker or web UI.
Updated 4 months ago (2026-07-26 14:15 UTC)
5
-
coord-mem MCP Server
Knowledge & Memory Autonomous Agents
khalildh
A
license
Not graded
quality
D
maintenance
Memory-palace graph MCP server for AI agent memory, exposing memory_add / get / like / record_window / walk / retrieve / neighbors / rooms / misshelved / rehome operations over a zero-dependency, append-only, provenance-tracking knowledge graph with decay, PMI-derived adjacency, and spatial coordinates.
Updated 3 months ago (2026-08-03 18:44 UTC)
MIT
worldbrain-mcp
Autonomous Agents Knowledge & Memory
genesis-plan
A
license
Not graded
quality
F
maintenance
Enables AI agents to perform auditable A* reasoning, import world graphs, incorporate physical carrier constraints, and maintain a learning knowledge base through a zero-dependency local MCP stdio service.
Updated 3 months ago (2026-08-26 06:34 UTC)
MIT
Algentaofficial
Agent Orchestration AI & Machine Learning Autonomous Agents
thyn-ai
A
license
A
quality
A
maintenance
Official MCP server for Algenta, a local-first AI engine powered by Mojo CPU/GPU kernels, 525+ Mojo libraries and 6,000+ compiled functions. Governed data discovery, exact queries, decisions, simulations, persistent decision memory and agent runs with approvals and execution receipts. Runs on your infrastructure. Source: packages/mcp/algenta_mcp. Install: pip install algenta-mcp (stdio).
Updated 2 days ago (2026-11-09 16:44 UTC)
140
1
Apache 2.0

Graph Attention Engine (GAE)

Install

Quick Start

Scaled dot-product attention

Profile scoring

Core API

Core Scoring

Oracle

Evaluation

Judgment

Calibration

Primitives

Learning

Convergence

Kernels

Persistence

Conservation Law

MCP Server

Install

Usage

Docker

MCP Tools

Claude Desktop Configuration

MCP Server

Install

Usage

Docker

MCP Tools

Claude Desktop Configuration

Testing

Documentation

License

Tool Schema Changelog

Maintenance

Resources

Related MCP Connectors

Related MCP Servers

intelligraph-mini

coord-mem MCP Server

worldbrain-mcp

Algentaofficial