Agent Skills: Generate scientific hypotheses

Generate testable scientific hypotheses from observations, data, or research questions. Develop competing explanations, design experiments, and formulate predictions. Use during the PLANNING phase or when developing research aims.

UncategorizedID: braselog/researchassistant/hypothesis-generation

Install this agent skill to your local

pnpm dlx add-skill https://github.com/braselog/researchAssistant/tree/HEAD/.github/skills/hypothesis-generation

Skill Files

Browse the full folder contents for hypothesis-generation.

Download Skill

Loading file tree…

.github/skills/hypothesis-generation/SKILL.md

Skill Metadata

Name
hypothesis-generation
Description
Develops competing, testable hypotheses and discriminating predictions from project observations and verified evidence. Use for /hypothesis_generation, research aims, unexpected findings, or experimental planning.

Generate scientific hypotheses

Read the relevant observation, design, project aims, prior analyses, and literature. Separate established evidence from assumptions.

Workflow

  1. Define the phenomenon, observational unit, scope, and uncertainty.
  2. Identify plausible mechanisms at relevant biological or technical levels.
  3. Generate multiple genuinely competing explanations, including confounding, measurement, and pipeline artifacts where appropriate.
  4. For each hypothesis, state supporting evidence, contradictory evidence, falsifying observations, and discriminating predictions.
  5. Propose the smallest informative test, required controls, expected outcomes, and alternative interpretations.
  6. Identify feasibility, identifiability, and ethical constraints.

Do not force a fixed number of hypotheses, rank novelty above validity, or convert hypotheses into project aims without user approval. Record an approved change to project direction through record-decision and update .research/project_telos.md only when requested.