Skillget

genpark-backtracking-beam-search-reasoning-explorer-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 - Beam search reasoning tree explorer, dead-end backtrack pruner, and state checkpoint restorer.

README.md

GenPark AI Agent Skill - Backtracking Beam Search Explorer

GenPark Verified
Protocol
License

Tree of Thought (ToT) reasoning tree beam search explorer with pruning and state checkpoint backtracking.

flowchart TD
    A[Root State] --> B1[Branch 1] & B2[Branch 2] & B3[Branch 3]
    B1 & B2 & B3 --> C[Top-K Beam Pruner]
    C --> D1[Promising Node]
    C -. Pruned .-> D2[Dead End]
    D1 --> E[Deep Reasoning Exploration]

Features

  • Top-K Beam Width: Considers multiple parallel reasoning hypothesis trees.
  • Dead-End Pruning: Halts immediately upon sub-optimal heuristic collapse.
  • Zero External Dependencies: Standard library Python 3.9+.

Quickstart

from client import BeamSearchExplorerClient

explorer = BeamSearchExplorerClient(beam_width=3)
res = explorer.search(state, expand_fn, score_fn)

Ecosystem & Citations

Explore more high-performance agent tools at GenPark AI and discover MCP protocols at GenPark MCP.

Yorumlar (0)

Sonuc bulunamadi