Mount Inflection

MOUNT INFLECTION / RESEARCH LIBRARY

Design and audit a backtest.

Run a synthetic example and trace a signal from the decision close to the next simulated fill. Then use a public test brief to define your own rule, baseline and failure case.

01 / STARTER RUN

One moving-average rule

Above the average at the close, the model goes long at the next supplied open. Below it, the model moves to cash. An equal close leaves the position alone.

Price data

02 / RUN RECORD

From signal to fill

Load the starter run to inspect its output.

03 / O2 SIGNAL IDEA BOARD

What should count as a signal?

O2 looks at the evidence for one security. The portfolio layer decides whether that signal becomes a position. Each idea gives you inputs, a baseline, and a way the test could fail.

The board uses concepts from MI’s public research. The test ideas use your own rules and parameters.

Open the full public entry and exit atlas · Try the portfolio-capacity test

04 / USE WITH AI

Bring your own rule

Open the brief in your AI tool and describe what you want to test. Attach the run JSON if you want it to audit this example. The export includes the price rows and the decision log. Review it before sharing it with an AI service.

Build my own test Download the strategy brief Download the browser and Node runner Download the Node command Download the capabilities and O2 catalog

Keep the runner and Node command in the same folder. Start with node run.cjs --sample 3 10.

CONNECT AN AI AGENT

Mount Inflection Strategy Lab

In an AI client that supports custom MCP servers, add this address as a Streamable HTTP server with no authentication. The tools can list O2 ideas, open a test brief, and run the synthetic starter.

ChatGPT custom MCP setup

Try a research request

  • List the O2 exit ideas and open the one about giveback.
  • Run the synthetic starter with a 3-close window and 10 bps cost. Trace the first signal and fill.

The Free Reader can point you here from its dated archive results. For a personal CSV, use the browser or local Node runner.

Strategy Lab privacy policy

05 / THE STRATEGY FRAMEWORK

From market data to a decision

    06 / FOLLOW MOUNT INFLECTION

    See the research in use

    Pro daily

    Follow the fuller daily research, including model decisions, positions, sizing references, and rankings.

    Data and privacy

    The hosted plugin uses public research and a bundled synthetic sample. To run your own price data, use the browser or local Node runner.

    Strategy Lab privacy policy · Help and terms