quantcoder-plugin AGENTS.md

Repository instructions for a QuantCoder plugin, which helps coding agents research quantitative papers and turn them into QuantConnect LEAN Python drafts. They also describe the plugin’s files, safety rules, and development checks.

In plain words
What is it for?
Use it when developing or maintaining the QuantCoder plugin, especially when changing its commands, skills, manifests, or MCP connection.
Why use it?
It gives agents one shared set of rules, reducing inconsistent behavior across Codex and Claude Code and helping keep plugin changes safe and aligned.

Instructions file for CodexOpenCode

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 instructions/nutdnuy/quantcoder-plugin/agents-md
Clone the repo
git clone --depth 1 https://github.com/nutdnuy/quantcoder-plugin

Made for: Codex, OpenCode.

Per session 423 This file is loaded in full into every session.
When invoked 423 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 $0.00423 $0.00423
Opus 5 $0.00211 $0.00211
Sonnet 5 $0.00085 $0.00085
Haiku 4.5 $0.00042 $0.00042

Measured yesterday against content hash 18d9a4ebe1e4, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

quantcoder-plugin AGENTS.md 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 yesterday.

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.

AGENTS.md · 56 lines

What it actually says

Agent Instructions - QuantCoder Plugin

This repository is the shared source of truth for the QuantCoder Claude Code and Codex plugin wrapper around SL-Mar/quantcoder.

Goal

Provide an installable agent plugin that helps users run QuantCoder workflows:

  • search quantitative papers
  • download and summarize papers
  • generate QuantConnect LEAN Python drafts
  • locally lint common QuantConnect API mistakes
  • optionally validate and backtest through QuantConnect
  • run evolution/autonomous/library-builder modes when appropriate

Plugin Surfaces

  • Codex plugin manifest: .codex-plugin/plugin.json
  • Codex marketplace manifest: .agents/plugins/marketplace.json
  • Claude plugin manifest: .claude-plugin/plugin.json
  • Claude marketplace manifest: .claude-plugin/marketplace.json
  • Shared skill: skills/quantcoder-research/SKILL.md
  • Claude slash commands: commands/*.md
  • MCP configuration: .mcp.json
  • MCP stdio wrapper: quantcoder/mcp_stdio.py and scripts/quantcoder-mcp

When behavior changes, update all relevant surfaces.

Development

Use uv when available:

uv run quantcoder --help
uv run --extra mcp python -c "import quantcoder.mcp_stdio; print('ok')"
uv run --extra mcp pytest tests/test_qc_linter.py tests/test_mcp.py -q

Validate the Codex plugin manifest before handoff:

python3 ~/.codex/skills/.system/plugin-creator/scripts/validate_plugin.py .

Safety

QuantCoder output is research scaffolding. Never present generated strategies or backtests as guaranteed profitable. Always distinguish implementation facts, assumptions, interpretations, and recommendations. Check for data leakage, look-ahead bias, survivorship bias, overfitting, unrealistic costs, and mathematical fidelity to the source paper.

Do not commit secrets. QuantConnect and optional API keys must come from user environment variables or ~/.quantcoder/.env.

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. yesterday First seen · 56 lines · 423 tokens per session scan A 18d9a4ebe1e4

Subscribe to this mod's changes

quantcoder-plugin AGENTS.md is an instructions file published in the GitHub repository nutdnuy/quantcoder-plugin (1 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 423 tokens to every session, about $0.0021 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.