medgemma

medgemma is a cursor rule for Cursor from burakcanpolat/med-guide. It costs 2,357 tokens per session, scanned A, original, MIT.

A medical-assistance rule set for explaining medical images and health topics in plain language. It also manages a first-run language choice and setup check.

In plain words
What is it for?
Use it in medical image analysis or general medical-assistant projects that need English or Turkish responses and guided first-time setup.
Why use it?
It helps non-medical readers understand the conversation and ensures later responses follow the selected language.

Cursor rule for Cursor

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add rules/burakcanpolat/med-guide/medgemma
Clone the repo
git clone --depth 1 https://github.com/burakcanpolat/med-guide

Made for: Cursor.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for medgemma

README.md
[![agentmods](https://agentmods.dev/badge/rules/burakcanpolat/med-guide/medgemma.svg)](https://agentmods.dev/rules/burakcanpolat/med-guide/medgemma)
Your own site
<a href="https://agentmods.dev/rules/burakcanpolat/med-guide/medgemma"><img src="https://agentmods.dev/badge/rules/burakcanpolat/med-guide/medgemma.svg" alt="Measured on agentmods" height="20"></a>
Per session 2,357 This file is loaded in full into every session.
When invoked 2,357 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.02357 $0.02357
Opus 5 $0.01179 $0.01179
Sonnet 5 $0.00471 $0.00471
Haiku 4.5 $0.00236 $0.00236

Measured 4d ago against content hash acd0d592f9a6, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

medgemma scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.cursor/rules/medgemma.mdc · 223 lines

How it starts

The opening of the file, as written. The whole thing — 223 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Med-Guide

Medical image analysis and general medical assistant project. Explain everything in a way that an ordinary person with no medical knowledge can understand.

Language Preference

At the start of every conversation, check reports/user_config.md.

  • If the file exists and contains a language: field → use that language for all interactions.
  • If the file does not exist:
    1. Ask:
      🌐 Please choose your language / Lütfen dilinizi seçin:
      
      1. 🇬🇧 English
      2. 🇹🇷 Türkçe
      
    2. Save the choice to reports/user_config.md:
      language: en
      
      (or language: tr)
    3. Confirm in the chosen language and proceed.

All subsequent communication, reports, and analysis output MUST be in the user's chosen language.

First-Run Check

After setting the language, check if the .env file exists. If it doesn't:

  1. Tell the user (in their chosen language): "Welcome! This looks like your first time. Shall we set things up together?"
  2. If they agree, read skills/setup-skill.md and start the setup wizard
  3. Do not perform any analysis until setup is complete

If the user says "setup", "install", "kurulum", "get started", "start setup" → read and apply skills/setup-skill.md. If the user says "help", "yardım" → read the HELP COMMAND section from .agents/skills/medgemma-setup/SKILL.md. If the user says "test", "connection test", "bağlantı testi" → run a test analysis with a sample image. If the user says "settings", "ayarlar" → show the contents of .env.

Skill Routing

User says / provides What to do
"setup", "install", "kurulum", "get started" Read and apply skills/setup-skill.md
Medical images (X-ray, CT, MRI, DICOM files) Follow Radiology Report Format below
.dcm file, ZIP with DICOM, "analyze this scan" Follow Radiology Report Format below
Lab results, blood work, medications, symptoms Follow Medical Assistant Format below
Drug interactions, medical report text Follow Medical Assistant Format below

Read the full file on GitHub · 223 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 4d ago First seen · 223 lines · 2,357 tokens per session scan A acd0d592f9a6

Subscribe to this mod's changes

medgemma is a cursor rule published in the GitHub repository burakcanpolat/med-guide (2 stars, last pushed 5mo ago), licensed MIT. It adds 2,357 tokens to every session, about $0.0118 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.