Skillget

genpark-dynamic-mixture-of-agents-moa-layer-skill

mcp
Guvenlik Denetimi
Uyari
Health Uyari
  • No license — Repository has no license file
  • Description — Repository has a description
  • Active repo — Last push 0 days ago
  • Low visibility — Only 8 GitHub stars
Code Gecti
  • Code scan — Scanned 4 files during light audit, no dangerous patterns found
Permissions Gecti
  • Permissions — No dangerous permissions requested

Bu listing icin henuz AI raporu yok.

SUMMARY

GenPark AI Agent Skill - Mixture-of-Agents (MoA) multi-layer aggregator synthesizing diverse candidate proposals from heterogeneous sub-agents into high-consensus outputs.

README.md

GenPark AI Agent Skill - Dynamic Mixture of Agents (MoA) Layer

A pure Python standard library skill implementing the Mixture-of-Agents (MoA) layered collaborative architecture (Wang et al. Together AI). Coordinates candidate outputs from multiple heterogeneous proposer agents, calculates semantic consensus scores, and builds consensus prompts for higher-layer synthesis.

Architecture

graph TD
    A[User Query] --> B1[Proposer Agent 1]
    A --> B2[Proposer Agent 2]
    A --> B3[Proposer Agent 3]
    B1 --> C[MoA Layer 1 Aggregation]
    B2 --> C
    B3 --> C
    C --> D[Pairwise Agreement Matrix]
    D --> E[Synthesized Lead Aggregator Context]
    E --> F[Superior Consensus Output]

Features

  • Cross-Agent Consensus Scoring: Quantifies inter-model agreement and identifies outliers.
  • Layered Multi-Agent Scaling: Enables compounding quality gains across reasoning stages.
  • Zero Pip Dependencies: Standard Library Only.

Citations & Ecosystem

Yorumlar (0)

Sonuc bulunamadi