Skip to content
Iman edited this page Sep 8, 2026 · 4 revisions

Option Desk guide

Option Desk turns option-chain snapshots into Greeks, positioning charts, strategy comparisons, and a local dashboard. Local tools also support simulation, backtesting, and a paper ledger.

Start here

What you want Where to start What you get
See the results first Dashboard tour Screenshots and a guide to each panel.
Try it without market data First walkthrough A sample chain, calculations, and a dashboard on your computer.
Use your agent Agent workflows Installation paths and prompts for local and hosted tools.
Analyze your own chain Import your data Required fields, an example file, and error recovery.
Run a research workflow Examples Structures, two-expiry spreads, simulation, backtests, and paper positions.
Resolve an error Troubleshooting Checks for installation, data, charts, and model results.

Choose where the calculation runs

Local desk Hosted service Skills only
Requirements Python and the installed tools A supported browser or agent connection An agent that reads skills
Input Your files or an enabled provider The hosted SYNTH sample or a permitted upload Your question and available context
Calculates results Yes Yes, within the hosted tool set Requires separately connected tools
Simulation and backtests Yes, with price history Not part of the hosted workflow Requires local tools
Dashboard Local artifact viewer Separate hosted service No dashboard by itself

Install the desk, then complete the first walkthrough. The hosted and local SYNTH examples are separate teaching samples. Their numbers need not match.

Understand and maintain the desk

  • Read the results: units, missing data, model assumptions, and comparison limits.
  • Architecture: packages, artifacts, and the calculation path.
  • Master algorithm: skills, loops, graph routing, prompts, backtests, and paper tests in one pseudocode reference.
  • Development: checks, screenshots, and documentation maintenance.
  • Documentation map: guides and reference material.

Research software. No order placement. Results are not investment advice. The software is free for noncommercial use under the project license. Privacy and the research disclaimer explain the boundaries.

Next: Complete the first walkthrough.

Install skills by platform · Architecture and preserved reference

Clone this wiki locally