AI Agent Discipline Linter

Version updated for https://github.com/naimkatiman/continuous-improvement to version v3.1.0.

  • This action is used across all versions by 0 repositories.

Action Type

This is a Node action using Node version 20.

Go to the GitHub Marketplace to find the latest changes.

Action Summary

This GitHub Action implements “The 7 Laws of AI Agent Discipline” to enhance the reliability and effectiveness of AI agents. It addresses common issues like skipped steps, lack of verification, and repetitive mistakes by enforcing structured workflows and a learning system (Mulahazah) to build better instincts over time. The action automates disciplined task execution, reducing errors, improving efficiency, and ensuring more thoughtful decision-making by AI agents.

What’s Changed

What’s New

GitHub Action: Transcript Linter

  • CI/CD lint your AI agent transcripts for compliance with the 7 Laws of AI Agent Discipline
  • Runs as a GitHub Action in your workflow

MCP Server (v3.0.0+)

  • Zero-dependency JSON-RPC stdio server for instinct management
  • Beginner mode (3 tools) / Expert mode (8 tools)
  • Works with Claude Code, Claude Desktop, Cursor, Zed, Windsurf, VS Code

Starter Instinct Packs

  • react.json, python.json, go.json — pre-built instinct sets

Community & Ecosystem

  • llms.txt for LLM discoverability
  • X thread template in examples/
  • 34-test suite (up from 20)

Install

npx continuous-improvement install
Full changelog: https://github.com/naimkatiman/continuous-improvement/blob/main/CHANGELOG.md



**Steps:**
1. Click **"Select tag"** → pick `v3.1.0`
2. Paste the title and release notes above
3. Check **"Set as the latest release"** 
4. Click **"Publish release"**