360 Secure SkillHub
360 安全技能中心

Zouroboros Bench

面向 AI 记忆系统的基准评测框架：通过命令行参数运行 LongMemEval、LoCoMo、ConvoMem 等数据集，调用 zouroboros-memory 后端执行测试，可选 GPT-4o 或本地 Ollama 的 Mimir 评审检测架构漂移，并生成运行报告。

marlandoj
marlandoj
热度 38.08
最近更新 2026-05-18 11:44:48
分类：开发依赖：代码执行 / 读写文件 / 第三方 API / 内部调用大模型应用场景：报告撰写 / 数据处理 / 自动化脚本 / 工作流

安装技能

复制下面的话术发给你的 Agent，它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill： npx secure-skills-cli add zouroboros-bench
Metadata
name
zouroboros-bench
description
Benchmark harness for AI memory systems. Evaluates LongMemEval, LoCoMo, and ConvoMem datasets against any memory backend via the zouroboros-memory CLI. Includes Mimir judge for catching architectural drift.
version
1.0.0
compatibility
OpenClaw, Claude Code, Codex CLI, any Node.js 22+ environment
metadata
{"author":"marlandoj.zo.computer","openclaw":{"emoji":"📊","requires":{"bins":["node"],"env":["OPENAI_API_KEY"]},"primaryEnv":"OPENAI_API_KEY","install":[{"id":"node-zouroboros-bench","kind":"node","package":"zouroboros-bench","bins":["zouroboros-bench","zouroboros-bench-report"],"label":"Install Zouroboros Bench (npm)"}],"homepage":"https://github.com/AlaricHQ/zouroboros-openclaw"}}

Usage

Install: npm install zouroboros-bench zouroboros-memory

Run all benchmarks

npx zouroboros-bench --limit 50

Run specific benchmark

npx zouroboros-bench --benchmarks longmemeval --limit 100 --judge

Generate report

npx zouroboros-bench-report --runs ./data/runs/

Environment Variables

  • ZOUROBOROS_MEMORY_CLI — Path to memory CLI binary (default: zouroboros-memory)
  • ZOUROBOROS_MEMORY_DB — SQLite DB path for benchmarks
  • OPENAI_API_KEY — Required for GPT-4o judge
  • OLLAMA_URL — Ollama URL for local LLM (default: http://localhost:11434)