skills.rest
open-horizon-labs
open-horizon-labsOFFICIAL
Explore all skills by this author →
GitHub

sg

Evaluate AI decisions with sg CLI commands like review, init, and audit.

3|Updated Jan 7, 2026
Search Tags:#CLI tools#metacognition#project governance#ai oversight#superego#decision review
One-click install
npx skills add https://github.com/open-horizon-labs/bottle --skill sg
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sg
Source: https://github.com/open-horizon-labs/bottle/tree/main/codex-skill/sg
Command: npx skills add https://github.com/open-horizon-labs/bottle --skill sg

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Superego provides metacognitive review for AI sessions, catching premature commitment, scope creep, and misalignment.

Core Features & Use Cases

  • On-demand evaluation: Run structured metacognitive checks before plans are committed.
  • Decision history: Query decision history to understand why and how conclusions were reached.
  • Project grounding: Initialize and maintain project-specific cognitive safeguards.

Quick Start

Initialize superego for the project: sg init Perform an on-demand review: sg review Inspect decision history: sg audit

Frequently Asked Questions about sg

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
What is metacognitive oversight for AI decisions and when is it needed?▼

Metacognitive oversight for AI decisions evaluates sessions to catch premature commitment, scope creep, and misalignment. It is needed during planning, execution, and review phases to prevent drift and ensure alignment with project goals.

How do I run a metacognitive review of AI planning decisions?▼

To run a metacognitive review, execute the 'sg review' command in your environment. This performs an on-demand structured evaluation to validate decisions before plans are committed.

How do I initialize project governance safeguards for AI sessions?▼

Initialize project governance safeguards by running the 'sg init' command. This establishes project-specific cognitive safeguards to ground AI decisions and prevent scope creep throughout the session.

Can I audit the history of AI decisions to understand why a conclusion was reached?▼

Yes, you can audit AI decision history by running the 'sg audit' command. This queries past decisions to understand exactly why and how specific conclusions were reached during your session.

Do I need to install a specific CLI tool to enable AI oversight?▼

Yes, enabling AI oversight requires the 'sg' CLI tool to be installed. The CLI provides the review, init, and audit commands necessary to evaluate and validate decisions within your project.

Related Skills

Explore More Software Engineering Skills→
Cobalt-MoneyCobalt-Money
7

stripe-projects

Provision software stacks and initialize local repositories using the Stripe Projects CLI.

Official
Basic
Pratikkadam254Pratikkadam254

avoid-feature-creep

Validate feature requests against problem validity, alignment, impact, and complexity.

Community
Advanced
HaRu-LLCHaRu-LLC

overall-charter

Create an overarching program charter coordinating sub-projects under unified governance.

Community
Advanced
BellaBeBellaBe
37

reasoning-counterfactual

Evaluate alternative decisions by simulating interventions on past or hypothetical futures.

Community
Advanced