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.
npx agentmods add instructions/supportersimulator/3-surgeons/gemini-mdgit clone --depth 1 https://github.com/supportersimulator/3-surgeonsWrote 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.
[](https://agentmods.dev/instructions/supportersimulator/3-surgeons/gemini-md)<a href="https://agentmods.dev/instructions/supportersimulator/3-surgeons/gemini-md"><img src="https://agentmods.dev/badge/instructions/supportersimulator/3-surgeons/gemini-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.01037 | $0.01037 |
| Opus 5 | $0.00518 | $0.00518 |
| Sonnet 5 | $0.00207 | $0.00207 |
| Haiku 4.5 | $0.00104 | $0.00104 |
Grade B, and why
3-surgeons GEMINI.md scanned grade B with 1 finding 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 5d 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
If the extension install does not wire MCP automatically, add this to your Gemini settings (`~/.gemini/settings.json`): How it starts
The opening of the file, as written. The whole thing — 122 lines — stays where its author put it; the contents beside it link to each section on GitHub.
3-Surgeons for Gemini
Three AI surgeons. One operating table. Your code doesn't ship until all three agree.
Setup
1. Install the extension
gemini extensions install https://github.com/supportersimulator/3-surgeons
2. Install Python runtime
Python 3.10+ is required. The MCP server auto-bootstraps on first launch, but you can install manually:
cd ~/.gemini/extensions/3-surgeons # or wherever Gemini places extensions
python3 -m venv .venv
.venv/bin/pip install -e '.[mcp]'
3. Configure surgeon backends
Copy a preset to your home config:
# Hybrid (default) — OpenAI cardiologist + local Ollama neurologist
cp config/presets/hybrid.yaml ~/.3surgeons/config.yaml
# API-only — OpenAI + DeepSeek, no local model needed
cp config/presets/api-only.yaml ~/.3surgeons/config.yaml
# Local-only — all Ollama, $0 cost, fully private
cp config/presets/local-only.yaml ~/.3surgeons/config.yaml
4. Set API keys (if using API-backed surgeons)
# For hybrid or API-only mode
export OPENAI_API_KEY="sk-..."
# For API-only mode (DeepSeek neurologist)
export DEEPSEEK_API_KEY="sk-..."
Add these to your shell profile (~/.zshrc, ~/.bashrc) so they persist across sessions.
5. Verify
In Gemini, run the probe tool to confirm all three surgeons are reachable:
Use the probe tool to check surgeon connectivity
Expected: all three surgeons report OK with model names and response times.
Alternative: Manual MCP configuration
If the extension install does not wire MCP automatically, add this to your Gemini settings (~/.gemini/settings.json):
{
"mcpServers": {
"3-surgeons": {
"command": "/ABSOLUTE/PATH/TO/3-surgeons/bin/3surgeons-mcp",
"args": []
}
}
}
Replace /ABSOLUTE/PATH/TO/3-surgeons with the actual path where you cloned the repo.
Available tools
Once connected, these MCP tools are available in Gemini:
| Tool | What it does |
|---|---|
probe |
Verify all three surgeons are reachable |
cross_examine_start |
Begin a 3-phase cross-examination on a claim |
cross_examine_deepen |
Phase 2: surgeons challenge each other's findings |
cross_examine_synthesize |
Phase 3: final synthesis with confidence scores |
consensus |
Quick confidence-weighted vote from multiple surgeons |
sentinel_run |
Scan for complexity risks across configurable dimensions |
ask_local_tool |
Query the local neurologist directly |
ask_remote_tool |
Query the remote cardiologist directly |
research_tool |
Self-directed multi-surgeon research on a topic |
cardio_review_tool |
External-model cross-examination review |
neurologist_pulse_tool |
Local model health check |
neurologist_challenge_tool |
Corrigibility challenge from the neurologist |
introspect_tool |
Surgeon capability introspection |
gains_gate |
System health and safety invariant checks |
ab_propose |
Propose an A/B test |
ab_start |
Start a proposed A/B test |
ab_measure |
Measure A/B test results |
ab_conclude |
Conclude an A/B test with surgeon consensus |
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.
- 5d ago First seen · 122 lines · 1,037 tokens per session scan B 914cd898b92f
3-surgeons GEMINI.md is an instructions file published in the GitHub repository supportersimulator/3-surgeons (2 stars, last pushed 3mo ago), licensed MIT. It adds 1,037 tokens to every session, about $0.0052 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
academic-commercialization-agent AGENTS.md
AGENTS.md instructions for shuxiachai/academic-commercialization-agent, covering agents.md — working on this project, what it is, commands, hard conventions and where the reasoning lives.
Tianshu-harness CLAUDE.md
Claude Code instructions for huiliyi37/Tianshu-harness, covering 天枢 (tianshu) / rivet, build & test, architecture, conventions and known constraints.
cherry-studio AGENTS.md
AGENTS.md instructions for CherryHQ/cherry-studio, a project described as: AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs.
rikkahub AGENTS.md
AGENTS.md instructions for rikkahub/rikkahub, covering repository guidelines, project overview, build, test, and development commands, module structure and concepts.
pcl copilot-instructions.md
Copilot instructions for personamanagmentlayer/pcl, covering project overview, pcl bootstrap system, core architecture principles, 1. compiler pipeline and 2. type system.
seektty AGENTS.md
AGENTS.md instructions for Hilbert-beinghappy/seektty: This repository ships one out-of-tree DeepSeek Harness Bundle. Harness remains the only owner of Agent, Session, model, settings, permissions, Profile, plugin, and persistence state.