Calculator

Stdin-JSON add/subtract/multiply/divide script that Claude beats on its own

Tested · Didn't pass

Test report

Verdict
Tested · Didn't pass
Tested
Jul 17, 2026
Environment
Claude Code 2.x (agent harness)
Upstream re-checked
Aug 10, 2026 · fbbd956

Ran the bundled script head-to-head against a no-skill baseline and the skill lost on correctness: it coerces every input to float, so adding 1 to 12345678901234567890 returned 1.2345678901234567e+19 -- the +1 silently vanished -- while the baseline (plain Python, what Claude does anyway) returned the exact 12345678901234567891. Simple cases like 847*293 matched the baseline but arrived wrapped in float noise (248171.0). The skill adds a subprocess round-trip to operations Claude already performs exactly, and its only failure-handling win (division-by-zero JSON error) doesn't offset the precision regression. Install is clean and the docs honestly describe a simple calculator, but a skill that can produce wrong arithmetic is worse than no skill.

Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works

  • Installs cleanly 5/5
  • Triggers reliably 4/5
  • Output vs. baseline 3/10
  • Docs & honesty 4/5

What Calculator does

A minimal calculator skill: a Python script that reads a JSON operation from stdin and returns add, subtract, multiply, or divide results. Extracted from the evotown agent-arena repo. Triggers on basic arithmetic requests, where Claude's own computation is already exact and faster.

How to install Calculator

git clone https://github.com/EXboys/evotown
mkdir -p ~/.claude/skills
cd evotown && cp -r backend/arena_skills/calculator ~/.claude/skills/calculator

Skills live in ~/.claude/skills/ (global) or .claude/skills/ (per-project). Restart Claude Code after installing.

Commands — how to trigger Calculator

  • /calculator Stdin-JSON add/subtract/multiply/divide script that Claude beats on its own

It also activates on plain-language prompts like these:

  • Can you multiply 847 by 293 for me and give the exact number?
  • I need 12345678901234567890 plus 1, and it has to be exact
  • What's 20 divided by 4? Just double-checking the math here

Frequently asked questions

Is the Calculator skill free?
Yes. The skill itself is free from EXboys/evotown. SkillProof publishes the install command and an independent test verdict at no cost.
Does Calculator work with Claude Code?
We tested it with Claude Code 2.x (agent harness) on Jul 17, 2026. Verdict: Tested · Didn't pass. Ran the bundled script head-to-head against a no-skill baseline and the skill lost on correctness: it coerces every input to float, so adding 1 to 12345678901234567890 returned 1.2345678901234567e+19 -- the +1 silently vanished -- while the baseline (plain Python, what Claude does anyway) returned the exact 12345678901234567891. Simple cases like 847*293 matched the baseline but arrived wrapped in float noise (248171.0). The skill adds a subprocess round-trip to operations Claude already performs exactly, and its only failure-handling win (division-by-zero JSON error) doesn't offset the precision regression. Install is clean and the docs honestly describe a simple calculator, but a skill that can produce wrong arithmetic is worse than no skill.
How do I install Calculator?
Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
Can I use Calculator with Cursor, Copilot, Gemini CLI, Codex or other AI tools?
The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.