Exploration & Discovery

In the Exploration pattern, the goal is not to execute a known task, but to find new information or solutions. Agents act as scientists or researchers: they formulate a hypothesis, test it (by searching, coding, or simulating), analyze the results, and iterate. This allows for genuine novelty and discovery.

When to Use

• Literature Review: "Survey the field of Quantum Computing and identify gaps."
• Idea Generation: "Brainstorm 50 potential names for this product and check domain availability."
• Scientific Discovery: Analyzing large datasets to find correlations.
• Market Research: Exploring competitor websites to map out their feature sets.

Use Cases

• Agent Laboratory: A team of agents (Professor, Postdoc, Reviewer) writing a research paper.
• Creative Studio: Agents collaborating to write a screenplay or design a game.
• Scenario Planning: Simulating how a stock portfolio would perform under various economic conditions.

Implementation Pattern