analysis-poly

Real Profit (Fee-Excluded View)

中文说明见：README_zh.md

Polymarket market PnL analyzer with a web UI.

Scope (Important)

• The dashboard discovers markets from Polymarket user activity within the selected time range, then replays the matched markets.
• It is intended for Polymarket market analysis in general, including but not limited to crypto markets.
• Main purpose: quantify and visualize realized PnL, fee impact, market-level performance, and Net PnL vs No-Fee PnL.

Requirements

• Python 3.11+ (recommended: 3.11)
• uv package manager is optional for local development

Quick Start

pip install analysis-poly
analysis-poly-server --host 127.0.0.1 --port 8000

Open http://localhost:8000.

Local Install From Source

uv pip install .
# or
pip install .

Then start the web server with the same one-line command:

analysis-poly-server --host 127.0.0.1 --port 8000

For source checkout development, this also works:

uv sync
uv run python main.py

CLI Open + Auto Start

Use a standalone script to start server, open browser, and pass params in URL.

uv run -m analysis_poly.open_with_params \
  --address 0xabc \
  --keywords updown,15m \
  --start-time "2026-03-01 00:00" \
  --end-time "2026-03-02 00:00" \
  --concurrency 8

Frontend will read query params, fill form fields, and auto start the run.

CLI Run Once

Use the synchronous CLI entry when you want a single final JSON payload on stdout without SSE or web UI:

analysis-poly-run \
  --address 0xabc \
  --keywords updown,15m \
  --start-time "2026-03-01 00:00" \
  --end-time "2026-03-02 00:00"

The command blocks until analysis finishes, writes report artifacts into output_dir, and prints:

{
  "status": "COMPLETED",
  "summary": { "...": "..." },
  "artifacts": {
    "json": "reports/...",
    "total_curve_csv": "reports/...",
    "market_curve_csv": "reports/..."
  }
}

First Clone

analysis_poly/static/dist is committed in the repository, so first startup does not require front-end build.

If you modify frontend/src, rebuild assets:

npm install
npm run build

API

• POST /api/runs
• GET /api/runs/{run_id}/stream (SSE)
• POST /api/runs/{run_id}/stop
• GET /api/runs/{run_id}/result
• GET /api/runs/{run_id}/state

Test

uv run pytest

Frontend

• Source: frontend/src
• Build output: analysis_poly/static/dist/app.js and analysis_poly/static/dist/app.css