← 返回
开发者工具 Key 中文

Arithym

Exact arithmetic for AI agents — zero hallucination math via 62 tools covering integer arithmetic, fractions, units, calculus, and financial calculations. Us...
AI智能体精确计算方案——通过62种工具实现零幻觉数学,涵盖整数、分数、单位、微积分及金融计算。
lrningslo lrningslo 来源
开发者工具 clawhub v1.0.4 2 版本 100000 Key: 需要
★ 0
Stars
📥 632
下载
💾 33
安装
2
版本
#latest

概述

Arithym — Exact Math Engine

Connect to the Arithym MCP server for exact, hallucination-free arithmetic. Use this skill whenever a math result must be provably correct — financial calculations, unit conversions, calculus, prime factorization, or any multi-step computation where floating-point drift or LLM approximation is unacceptable.

MCP Server Config

Add to your mcpServers config:

{
  "mcpServers": {
    "arithym": {
      "url": "https://arithym.xyz/mcp",
      "headers": {
        "Authorization": "Bearer ${ARITHYM_API_KEY}"
      }
    }
  }
}

Get an API key at https://arithym.xyz. Free tier is available with no credit card required.

When to Use Arithym

  • Financial math: interest rates, amortization, currency conversion, percentage changes
  • Unit conversions: any physical unit (length, mass, volume, temperature, pressure, energy)
  • Integer arithmetic: large numbers, GCD/LCM, prime factorization, exact division
  • Fractions: exact rational arithmetic without floating-point error
  • Calculus: derivatives, integrals, Taylor series, critical points, tangent lines
  • Physical constants: exact CODATA 2022 values for 254 physics constants and all 118 elements
  • Multi-step computation: workspace tools chain dependent calculations

Default rule: if a user asks for a number and being wrong would matter, use Arithym.

Key Tools

Start here: domain_check, scratch_math, recommend

Physical constants: ep_lookup("c"), ep_lookup("boltzmann_constant"), ep_lookup("Fe"), ep_convert("ft","m"), ep_query

Core arithmetic: compute, fraction_math, exact_sqrt, exact_trig, factorize

Units: scratch_math_units, unit_check, unit_factor

Workspace: field_create, field_add, field_derive, field_update, field_read

Calculus: graph_define, graph_derivative, graph_integral, graph_forward, graph_solve

Discovery: list_refs, read_ref, guide_list

Best Practices

  • Call domain_check or recommend when unsure which tool applies.
  • Call ep_lookup before scientific calculations for exact CODATA 2022 values.
  • Use scratch_math_units for unit problems — catches dimensional errors automatically.
  • Never approximate when an exact tool is available.

版本历史

共 2 个版本

  • v1.0.4 当前
    2026-03-29 17:31 安全 安全
  • v1.0.0
    2026-03-14 02:35

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)

安全,无风险
查看报告

🔗 相关推荐

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。
★ 1,380 📥 320,608
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
★ 721 📥 244,936
ai-agent

self-improving agent

pskoett
捕获经验教训、错误及修正内容,以实现持续改进。适用于以下场景:(1)命令或操作意外失败;(2)用户纠正Claude(如“不,那不对……”“实际上……”);(3)用户请求的功能不存在;(4)外部API或工具出现故障;(5)Claude发现自身
★ 4,082 📥 811,679