Can I use Superforecasting for technical project management?

Yes, it is highly effective for estimating the probability of project delays, the success of a technology migration, or the impact of a code refactor on performance.

What is Superforecasting in the context of Claude Code?

Superforecasting is a specialized skill that teaches Claude to apply Philip Tetlock’s methodology for making predictions. It focuses on breaking down problems, using base rates, and assigning granular probabilities to outcomes.

How does this skill improve decision-making?

It eliminates common cognitive biases like the 'inside view' trap and overconfidence by forcing a systematic analysis of conflicting causal forces and historical data.

What is the 'Outside View' mentioned in the skill?

The 'Outside View' involves looking at the base rate of similar past events to establish a starting probability before adjusting for the specific details of the current situation.

Superforecasting Thinking Pattern

Name: Superforecasting Thinking Pattern
Author: lev-os

bylev-os

0•

Productivity & Workflow

Implements Philip Tetlock’s systematic methodology for making accurate, probabilistic predictions and strategic risk assessments.

This skill integrates the Superforecasting framework into Claude’s decision-making process, allowing users to move beyond binary thinking toward rigorous, data-driven probabilistic models. By applying techniques like Fermi-style decomposition, base rate analysis, and incremental Bayesian belief updating, it helps quantify uncertainty in market entry, product launches, and resource allocation. It is an essential tool for strategic planning where traditional expert intuition often falls short of systematic, calibrated analysis.

Key Features

01Fermi-style problem decomposition to break down complex forecasts into researchable parts

02Base rate analysis and outside-view anchoring to prevent over-optimism

03Incremental Bayesian belief updating for reacting appropriately to new information

04Pre-mortem and post-mortem protocols to identify and learn from predictive errors

050 GitHub stars

06Granular probability estimation techniques to move beyond vague language

Use Cases

01Predicting product-market fit and adoption rates for new technology launches

02Quantifying strategic risks in competitive intelligence and market shifts

03Assessing the likelihood of meeting project milestones and engineering deadlines

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add lev-os/agents superforecasting

For use in Claude.ai and ChatGPT