A professional decision architecture for AI agents to classify, analyze, and document complex choices using structured frameworks.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install afrexai-decision-engine
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install afrexai-decision-engine using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Decision Engine?
The Decision Engine is a comprehensive framework designed to help decision architects make better choices through structured analysis and bias reduction. It provides a technical methodology for classifying decisions as irreversible one-way doors or reversible two-way doors, ensuring that speed and quality are always balanced. This tool is a vital part of the Openclaw Skills library for those looking to build organizational decision-making muscle using data-driven logic.
By leveraging tools like Weighted Decision Matrices and Second-Order Thinking, the engine allows developers and teams to move past gut feelings and into rigorous information gathering. It emphasizes the 70% rule, where decisions are made when enough information is available to act without unnecessary delay, making it an essential protocol for high-performance agentic workflows.
Decision Engine Use Cases
- High-stakes strategic planning involving multiple stakeholders and irreversible consequences.
- Technical vendor selection and comparison using weighted scoring criteria.
- Resource prioritization and roadmap planning using the RICE framework.
- Reducing organizational friction by establishing automated decision policies for recurring tasks.
- Performing pre-mortem risk analysis to identify and mitigate foreseeable project failures.
How Decision Engine Works
- Classification: Identify if the decision is Type 1 (high stakes) or Type 2 (reversible) to determine the required analysis depth.
- Information Gathering: Perform a 70% rule audit and an assumption test to validate high-impact beliefs.
- Framework Library: Apply a logic model such as Inversion, Regret Minimization, or Bayesian Updating depending on the specific context.
- Bias Mitigation: Run a 15-point cognitive bias checklist to ensure the decision isn't skewed by mental shortcuts like sunk cost or confirmation bias.
- Execution and Documentation: Assign roles via the RAPID framework and log the final choice in a structured Decision Record for future review.
Decision Engine Setup
To deploy this skill, ensure your agent environment is configured to parse the provided logic templates. You can install the framework directly into your workspace to enhance your Openclaw Skills capabilities.
# Initialize the Decision Engine workspace
mkdir -p ./agent/skills/decision-engine
# Register the natural language triggers
openclaw register-skill decision-engine --path ./skills/decision-engine.md
Configure your persistent storage to maintain a decision_log.md file to track long-term hit rates and outcomes.
Decision Engine Data Schema & Taxonomy
The Decision Engine organizes data using structured YAML blocks within Markdown files to ensure portability across the Openclaw Skills ecosystem.
| Schema Component | Purpose | Data Type |
|---|---|---|
| Decision Brief | Defines context, constraints, and success criteria | YAML Object |
| Weighted Matrix | Compares options against weighted scores | YAML/Table |
| Kill Criteria | Defines stop/pivot thresholds for project health | List |
| Decision Record | Historical log of the decision, context, and lessons | ADR Format |
Decision Engine Advanced Features
- Multi-Scenario Planning: Build Bull, Base, and Bear case projections with probability-weighted outcome calculations.
- Disagree-and-Commit Protocol: Structured social contracts for team execution after a final call is made by the decider.
- Automated Policy Generation: Convert recurring Type 3 decisions into codified rules to reduce cognitive load and meeting fatigue.
- 100-Point Quality Rubric: A quantitative method for scoring the integrity of the decision process itself, independent of the outcome.
- Risk-Adjusted Expected Value (EV): Calculate the mathematical expected value of different options while factoring in worst-case downside protection.
SKILL.md
Loading
Related Openclaw Skills
Debt Collection & Recovery Playbook
A professional toolkit for generating compliant debt recovery strategies, letter sequences, and payment plan frameworks.
Deal Desk AI Agent Skill
A structured framework for AI agents to automate non-standard deal reviews, discount approvals, and margin analysis.
Davis-Bacon & Prevailing Wage Compliance Agent
A comprehensive federal construction compliance agent for managing Davis-Bacon Act requirements, DOL Wage Determinations, and certified payroll workflows.
Database Engineering Mastery
A comprehensive framework for database design, optimization, migration, and operations across PostgreSQL, MySQL, and SQLite.
Demand Forecasting Framework
A sophisticated analytical framework for generating accurate demand predictions using a blend of statistical time series, causal regression, and qualitative methodologies.
Dental Practice Operations
A comprehensive operational framework for Openclaw Skills that helps dental offices maximize production, maintain compliance, and reduce no-shows.
Featured*
