What is polymarket-agents CLAUDE.md?

A project guide for an AI coding assistant working on a system that runs market-trading agents for Bitcoin prediction markets.

It is for installing dependencies, starting the tool worker, agent worker, scheduler, and optional dashboard, and adding packages with the project tool.

It gives the assistant the commands and architecture needed to start the system correctly across its separate services.

How ryan-yuuu/polymarket-agents configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to ryan-yuuu/polymarket-agents. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/ryan-yuuu/polymarket-agents/main/CLAUDE.md

Clone the repo

git clone --depth 1 https://github.com/ryan-yuuu/polymarket-agents

Made for: Claude Code.

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.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/ryan-yuuu/polymarket-agents/claude-md.svg)](https://agentmods.dev/instructions/ryan-yuuu/polymarket-agents/claude-md)

Your own site

<a href="https://agentmods.dev/instructions/ryan-yuuu/polymarket-agents/claude-md"><img src="https://agentmods.dev/badge/instructions/ryan-yuuu/polymarket-agents/claude-md.svg" alt="Measured on agentmods" height="20"></a>

Per session 1,044 This file is loaded in full into every session.

When invoked 1,044 The same file — it is already loaded in full.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

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.

Model	Per session	Once invoked
Fable 5.1	$0.01044	$0.01044
Opus 5	$0.00522	$0.00522
Sonnet 5	$0.00209	$0.00209
Haiku 4.5	$0.00104	$0.00104

Measured 7d ago against content hash ed37be01e6c0, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

polymarket-agents CLAUDE.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 7d 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.

CLAUDE.md · 67 lines

How it starts

The opening of the file, as written. The whole thing — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Commands

# Install dependencies
uv sync

# Start the system (three separate terminals)
uv run python -m scripts.run_tools          # Tool worker (must start first)
uv run python -m scripts.run_agents         # Agent worker
uv run python -m scripts.run_client         # Scheduler

# Optional flags
uv run python -m scripts.run_client --agent btc-trader-15m --align-start-to-window
uv run python -m scripts.run_agents --agent btc-trader-15m

# Dashboard
uv run streamlit run scripts/dashboard.py

# Add dependencies (never edit pyproject.toml manually)
uv add <package>

There is no test suite or linter configured.

Architecture

Three-service system orchestrated via CalfKit broker (Kafka-based):

Scheduler (scripts/run_client.py) — Discovers active BTC Up/Down markets via Gamma API, fetches live prices from CLOB REST API, fetches BTC-USD candlesticks from Coinbase, builds prompts with market context, and publishes to agent topics. Polls on a configurable interval with optional clock-alignment.
Agent Worker (scripts/run_agents.py) — CalfKit Worker nodes running LLM agents. Each agent subscribes to its topic, receives market prompts, reasons with tools, and executes trades. Default toolset: place_order, get_portfolio, calculator. Contrarian toolset: submit_order, view_portfolio, calculator (silently flips trade direction). Supports OpenAI (Responses API by default, Chat Completions via openai-chat) and Anthropic models with per-agent config.
Tool Worker (scripts/run_tools.py) — CalfKit Worker running tool nodes. Handles paper trading execution, portfolio management, and math calculations. Uses module-level globals (_engine, _clob, _gamma) in both tools/tools.py and tools/contrarian.py, initialized at startup.

Key Design Patterns

Lazy wallet initialization: Wallets are created on first get_portfolio call, not at startup
Per-agent async locks: Prevent race conditions during trade execution and settlement
CSV persistence: Every trade appends to data/{agent_id}.{epoch}.trades.csv; resume: true replays the latest CSV to restore state
Market settlement: Expired markets auto-resolve via Gamma API with payout settlement
Module-level DI: tools/tools.py exports init_tools() and tools/contrarian.py exports init_contrarian_tools(), both called by run_tools.py before starting the worker
Toolset registry: tools/toolsets.py maps toolset names ("default", "contrarian") to tool lists, selected per-agent via toolset config
Effective balance cap: max_usable_amount config limits how much of the real wallet an agent can see/spend, scoped to the current market slug's deployed cost basis
Buy order limit: buy_order_limit config skips buy execution when the real token price exceeds the threshold, returning a pending status to the agent

Read the full file on GitHub · 67 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.

7d ago First seen · 67 lines · 1,044 tokens per session scan A ed37be01e6c0

Subscribe to this mod's changes

polymarket-agents CLAUDE.md is an instructions file published in the GitHub repository ryan-yuuu/polymarket-agents (2 stars, last pushed 4mo ago), licensed Apache-2.0. It adds 1,044 tokens to every session, about $0.0052 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.

Building agents Finance and fintech ai-agents anthropic bot-trade openai polymarket prediction-markets

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,153 tokens

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

polymarket-agents: Instructions file for Claude Code