← 返回
未分类 中文

Zouroboros Swarm Gate

Zero-cost task classifier (~2ms) that decides if a task needs multi-agent orchestration. 7 weighted signals, no API calls. Part of the Zouroboros ecosystem.
零成本任务分类器(约2ms),判断任务是否需要多智能体编排。7个加权信号,无需API调用。属于 Zouroboros 生态系统。
marlandoj marlandoj 来源
未分类 clawhub v1.0.0 1 版本 99740.9 Key: 无需
★ 0
Stars
📥 385
下载
💾 0
安装
1
版本
#latest

概述

Zouroboros Swarm Gate

Mechanical task complexity classifier that determines whether a task warrants multi-agent swarm orchestration. Runs in ~2ms with zero API cost — pure regex + heuristic scoring.

Usage

CLI

npx zouroboros-swarm-gate "build a REST API with auth, tests, and deploy to production"
# → [Swarm Decision Gate: SWARM — score 0.68]

npx zouroboros-swarm-gate --json "fix the typo on line 42"
# → { "decision": "DIRECT", "score": 0.05, ... }

Programmatic

import { evaluate } from "zouroboros-swarm-gate";

const result = evaluate("implement dashboard with 9 components, API routes, and deploy");
console.log(result.decision); // "SWARM"
console.log(result.score);    // 0.72
console.log(result.signals);  // { parallelism: 0.8, scopeBreadth: 0.6, ... }

Exit Codes

CodeDecisionMeaning
-------------------------
0SWARM / FORCE_SWARMTask warrants multi-agent orchestration
2DIRECTExecute directly, no orchestration needed
3SUGGESTOrchestration recommended but not required
1ErrorSomething went wrong

7 Signals

SignalWeightWhat It Detects
--------------------------------
parallelism20%Multiple independent workstreams, action verbs, lists
scopeBreadth15%Files, systems, and domains referenced
qualityGates15%Test/validation/review requirements
crossDomain15%Multiple executor types (code, design, ops, etc.)
deliverableComplexity15%Multiple output artifacts
mutationRisk10%Production/shared state changes
durationSignal10%Estimated effort and complexity

Part of the Zouroboros Ecosystem

Zouroboros is a self-improving AI orchestration framework. These standalone packages give you a taste of what's possible. For the full experience — persistent memory, swarm orchestration, scheduled agents, persona routing, and self-healing infrastructure — get a Zo Computer.

Built by @Xmarlandoj

版本历史

共 1 个版本

  • v1.0.0 当前
    2026-05-07 07:17 安全

安全检测

腾讯云安全 (Keen)

安全,无风险
查看报告

腾讯云安全 (Sanbu)

suspicious
查看报告

🔗 相关推荐

ai-agent

Skill Vetter

spclaudehome
AI智能体技能安全预审工具。安装ClawdHub、GitHub等来源技能前,检查风险信号、权限范围及可疑模式。
★ 1,233 📥 268,582
ai-agent

ontology

oswalpalash
类型化知识图谱,用于结构化智能体记忆与可组合技能。适用于以下场景:创建/查询实体(人物、项目、任务、事件、文档)、关联相关对象、强制执行约束、将多步操作规划为图谱变换,或当技能需要共享状态时。触发关键词包括"记住""我知道关于什么""将X链
★ 724 📥 245,280
ai-agent

Self-Improving + Proactive Agent

ivangdavila
自我反思+自我批评+自我学习+自组织记忆。智能体评估自身工作、发现错误并持续改进。
★ 1,387 📥 321,255