Hypothesis Tracker FAQs

Question 1

What is Hypothesis Tracker?

Accepted Answer

Hypothesis Tracker is an MCP (Multi-Modal Communication Protocol) server that provides AI agents with a persistent scientific method. It enables agents to create, track, and manage hypotheses, gather evidence, and update confidence levels using Bayesian logic across sessions.

Question 2

How does Hypothesis Tracker update hypothesis confidence?

Accepted Answer

It automatically updates hypothesis confidence using Bayesian logic. Confidence levels adjust based on whether new evidence is supporting, contradicting, or neutral, taking into account the weight or strength of the evidence. This provides a dynamic, data-driven assessment.

Question 3

Where is my data stored, and is it private?

Accepted Answer

All data generated by Hypothesis Tracker is stored locally on your machine in an SQLite database (specifically, `~/.hypothesis-tracker/data.db`). This ensures complete privacy and persistence, as no data leaves your machine.

Question 4

What are the primary use cases for Hypothesis Tracker?

Accepted Answer

Hypothesis Tracker is ideal for systematic debugging, evaluating architecture decisions, conducting root cause analysis, structuring research, and even performing detailed code reviews. It helps agents track competing theories and accumulate evidence for robust decision-making.

Question 5

Does Hypothesis Tracker provide an audit trail for hypotheses?

Accepted Answer

Yes, Hypothesis Tracker maintains a full audit trail and history for each hypothesis. You can access all evidence added, confidence changes, and final resolutions, ensuring transparency and traceability throughout the investigation process.

Hypothesis Tracker

Hypothesis Tracker

Key Features

Use Cases

Key Features

Use Cases