Decision Framework Skill for Openclaw

A structured decision-making system for AI agents that utilizes weighted confidence scores, redline interceptions, and historical memory to optimize choices.

kingofzhao
v1.0.0
Mar 31, 2026
0
683
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install kingofzhao-decision-framework

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install kingofzhao-decision-framework 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 Framework Skill?

The Decision Framework is a sophisticated Openclaw Skills tool designed to move AI beyond simple intuition toward structured, logical reasoning. It operates on the core philosophy that a good decision is defined by a structured process using known information to maximize expected value, rather than just the final outcome. By separating knowns from unknowns, it provides a transparent path for AI agents to evaluate complex trade-offs.

This skill is particularly powerful for developers and teams building autonomous agents that need to handle high-stakes environments. It incorporates a weighted scoring algorithm that prioritizes risk mitigation—penalizing potential downsides more heavily than rewarding upsides—to ensure safety and stability in automated workflows.

Decision Framework Skill Use Cases

  • Evaluating whether to adopt a new technology stack versus maintaining current stability.
  • Risk assessment for irreversible system changes or production deployments.
  • Automated business decision-making where confidence levels and financial trade-offs are involved.
  • Building self-evolving agents that learn from past decision outcomes through historical memory logs.

How Decision Framework Skill Works

  1. K (Known): The agent identifies all known facts and assigns a confidence level to each.
  2. U (Unknown): The agent lists uncertainties, assessing their potential impact and the cost required to acquire missing information.
  3. W (Weigh): The system applies the Decision Scoring Formula, calculating the Option Score by weighting known benefits/costs against probabilistic unknown upsides and downsides.
  4. R (Record): The decision, reasoning, and expected results are logged to a markdown file for future post-mortem analysis.

Decision Framework Skill Setup

Install the skill via the Openclaw Skills hub using the following command:

clawhub install decision-framework

Decision Framework Skill Data Schema & Taxonomy

The skill organizes decision data using a chronological file-based memory system stored in the decisions/ directory:

Feature Description
File Format {date}_{id}.md (Markdown)
Metadata Problem description, known/unknown lists, and weighted scores.
Post-Mortem Fields for actual results, deviation analysis, and lessons learned.
Redline Config Boolean flags and wait-timers for high-risk or irreversible decisions.

Decision Framework Skill Advanced Features

  • Asymmetric Risk Preference: Automatically weights unknown downsides (1.5x) more heavily than upsides (0.5x) to prevent catastrophic failures.
  • Redline Interception: Hard-coded safety checks that trigger 24-hour cooling-off periods for irreversible actions or mandatory consultations for social-impact decisions.
  • Decision Memory Evolution: A feedback loop system that updates decision parameters based on historical deviation analysis.
  • Confidence Thresholds: Automatically recommends information gathering if known info confidence falls below 50%.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*