dflow-mcp CLAUDE.md

Project instructions for dflow-mcp, an MCP server that lets AI clients access prediction-market metadata through tools. Prediction markets are services where people trade on the likelihood of future events.

In plain words
What is it for?
Use them when developing, building, testing, or running the dflow-mcp server with Bun, including its hot-reload development mode.
Why use it?
They give contributors one place to find the project’s commands, structure, API details, and development rules.

Instructions file

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/opensvm/dflow-mcp/claude-md
Clone the repo
git clone --depth 1 https://github.com/openSVM/dflow-mcp
Per session 1,092 This file is loaded in full into every session.
When invoked 1,092 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 $0.01092 $0.01092
Opus 5 $0.00546 $0.00546
Sonnet 5 $0.00218 $0.00218
Haiku 4.5 $0.00109 $0.00109

Measured 2d ago against content hash 62f1ff677184, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

dflow-mcp 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 2d 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 · 140 lines

How it starts

The opening of the file, as written. The whole thing — 140 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.

Project Overview

This is an MCP (Model Context Protocol) server that provides access to the Prediction Market Metadata API. The server wraps the REST API defined in llms_dflow.json and exposes it as MCP tools for use with Claude Desktop, Cursor, and other MCP clients.

Base API URL: https://prediction-markets-api.dflow.net

Commands

Development

# Start server in development mode with hot reload
bun run dev

# Start server in production mode
bun start

# Build the project
bun run build

# Run tests
bun run test

Running the server directly

bun run src/index.ts

Architecture

Single-File MCP Server

The entire server implementation is contained in src/index.ts. This is intentional for simplicity - the server is a thin wrapper around the external API.

Key components:

  1. DFlowAPIClient (lines 23-85): HTTP client that handles all API communication

    • Implements request timeout handling (30s default)
    • Constructs URLs with query parameters
    • Handles GET and POST requests
    • Returns parsed JSON responses
  2. TOOLS array (lines 103-696): Complete tool definitions

    • Each tool maps to one API endpoint
    • Tool definitions include JSON schemas for parameter validation
    • Organized by endpoint category (events, markets, trades, forecasts, candlesticks, live data, series, utilities)
  3. Request handlers (lines 698-849):

    • ListToolsRequestSchema: Returns all available tools
    • CallToolRequestSchema: Routes tool calls to appropriate API endpoints using a large switch statement
    • Error handling returns descriptive messages

API Endpoint Categories

The server exposes 25 tools across 8 categories:

  • Events: Get individual events, paginated event lists, search events
  • Markets: Get markets by ticker or mint address, batch queries
  • Trades: Query trade history with filtering and pagination
  • Forecasts: Historical forecast percentile data
  • Candlesticks: OHLC data for events and markets
  • Live Data: Real-time milestone information
  • Series: Series templates and metadata
  • Utilities: Outcome mints, filtering, tags, and search

Read the full file on GitHub · 140 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. 2d ago First seen · 140 lines · 1,092 tokens per session scan A 62f1ff677184

Subscribe to this mod's changes

dflow-mcp CLAUDE.md is an instructions file published in the GitHub repository openSVM/dflow-mcp (7 stars, last pushed 9mo ago), licensed Unlicense. It adds 1,092 tokens to every session, about $0.0055 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.