Foresight with Claude Code

A curated collection of projects and resources that combine Futures Studies methods with Claude Code.

This note focuses specifically on Claude Code as a tool, not on AI and foresight in general. The appeal lies in the combination of agentic coding, structured prompts, and established foresight methods.

Skills & Commands

Futurist Analyst Skill

URL: rysweet/MicrosoftHackathon2025

Core idea: Comprehensive futures methodology as a Claude Code skill.

Methods: Futures Cone, Scenario Planning, Three Horizons, Wind Tunneling, Delphi, Weak Signals

How it works: 9-step analysis process from focal question through STEEP scanning and scenario development to strategic options. Contains detailed frameworks for each step with explicit references to established methodologies.

Notable: Academically grounded with references to Voros, Shell Method. Very comprehensive: more of a reference work than a quick-start tool.

Future Scenario Generator

URL: buildwithclaude.com

Core idea: Slash command for systematic scenario generation with plausibility scoring.

Methods: STEEP/PESTLE, Cross-Impact Analysis, Morphological Analysis, Wild Cards

How it works:

  • Trend identification across multiple dimensions
  • Scenario portfolio generation (baseline, optimistic, pessimistic, transformation)
  • Plausibility scoring with weighted criteria (historical precedent, logical consistency, expert validation, empirical support)
  • Strategy testing against scenarios

Notable: Structured output format. Good balance between methodological rigor and practical usability.

Environmental Scanning & Foresight Skill

URL: MCP Hub

Core idea: Systematic environmental scanning with early warning capabilities.

Methods: PESTLE Analysis, Horizon Scanning, Weak Signal Detection, Signposts

How it works: Structured scanning across political, economic, social, technological, legal, and environmental dimensions. Includes patterns for specific contexts (industry disruption, regulatory change, climate/sustainability, geopolitical risk) and guidance for setting adaptive triggers.

Notable: Practical orientation with clear “when to use” guidance. Good for ongoing monitoring rather than one-off analysis.

MCP Servers

Chronulus MCP

URL: ChronulusAI/chronulus-mcp

Core idea: MCP server for AI-powered forecasting and prediction agents.

Methods: Quantitative Forecasting, Prediction Markets

How it works: Integrates with Claude Desktop and Claude.ai to provide forecasting capabilities. Focused on quantitative predictions rather than qualitative scenario development.

Notable: More suited for data-driven forecasting than strategic foresight. Consider as complement to scenario-based approaches.

Repositories & Frameworks

Shell Scenario Panel

URLs:

Core idea: Claude Code-powered scenario planning using the Shell method.

Methods: Shell Scenarios, Worldview Mapping, Lens-World-Lens Architecture

How it works: An 8-phase workflow orchestrated by a moderator AI (“Dr. Michelle Wells”):

  • Phase 0: Worldview elicitation: captures how you think about the topic before exploring scenarios
  • Phases 1-6: Classic Shell scenario development with 6 domain specialists (ecology, geopolitics, technology, economics, cultural anthropology, contrarian)
  • Phase 7: Worldview integration: connects scenarios back to your mental model

The Worldview Edition adds belief-by-belief analysis, crux mapping (connecting uncertainties to scenario boundaries), and personalized early warning signals.

Notable: Addresses a classic foresight challenge: scenarios that don’t connect to users‘ mental models. The fork extends this with structured belief tracking and quality gates.

Related: Future Scenario

VoltAgent Subagents Collection

URL: VoltAgent/awesome-claude-code-subagents

Core idea: Collection of specialized Claude Code subagents including foresight-relevant ones.

Methods: Trend Analysis, Scenario Planning, Competitive Intelligence

How it works: Provides ready-to-use subagent definitions for:

  • trend-analyst – emerging trends and forecasting
  • research-analyst – scenario planning, risk assessment, market research
  • competitive-analyst – competitive intelligence
  • market-researcher – market analysis and consumer insights

Notable: Building blocks rather than complete workflows. Good starting point for custom foresight agents.

Future-Forecasting-and-Prediction

URL: krisbolton/Future-Forecasting-and-Prediction

Core idea: Resource collection for scenario creation and future modelling.

Methods: Delphi Method, Scenario Creation, Future Modelling

How it works: Curated collection of PDFs and resources including foundational texts like Helmer’s “Analysis of the Future: The Delphi Method”.

Notable: Not Claude-specific, but valuable reference material for understanding the methodological foundations. Good for prompt engineering grounded in established methods.

Guides & Workflows

Using Claude for Geopolitical Scenario Development

URL: Memia Substack

Core idea: Concrete prompting workflow for accelerated scenario development.

Methods: Axes of Uncertainty, Quadrant Scenarios

How it works: System prompt positions Claude as strategic foresight analyst. User prompt feeds uncertainty axes and requests scenario generation with opportunities/threats for a specified entity. Author reports 10x acceleration compared to manual process.

Notable: Immediately applicable. Includes concrete prompt templates and real examples (geopolitical scenarios for New Zealand). Good starting point for adapting to own use cases.

Notes mentioning this note

Claude Code
“Documentation is no longer a record of what was built; it is the infrastructure through which we build.” — Matt...

Here are all the notes in this garden, along with their links, visualized as a graph.