Claudikins Kernel

We call it Claudikins because "Draconian-AI-Supervisor" was taken.

A disciplined workflow engine run by a team of neurotic AI agents.

Why?

You asked Claude for a bug fix. He refactored half your codebase.

You asked Claude for a feature. He placed a bunch of stubs that look a little bit real.

You asked Claude if you should drink that coffee you forgot about, now you're feeling sick and threw up on your keyboard, which cost too much just to hear nice click clack noises whilst you code, which actually isn't that great because it wakes your dog up at night, then you have to take him out in the cold to poop whilst you're still sick and well...okay, maybe that one was just me, but the point still stands!

claudikins-kernel applies SRE discipline to AI workflows. It enforces a strict 4-stage pipeline with gates between each step. You literally cannot skip verification. You cannot ship without the Cynic's approval.

Constraint is freedom. By preventing shortcuts, you get code that actually works.

The Workflow

┌──────────┐     ┌──────────┐     ┌──────────┐     ┌──────────┐
│ /outline │────▶│ /execute │────▶│ /verify  │────▶│  /ship   │
└──────────┘     └──────────┘     └──────────┘     └──────────┘
      │                │                │                │
      ▼                ▼                ▼                ▼
  taxonomy-        babyclaude      catastrophiser   git-perfectionist
  extremist        spec-reviewer       cynic
                   code-reviewer

Each arrow is a gate. Try to /ship without /verify passing? Blocked. Try to /execute without a plan? Blocked. The system enforces this - not guidelines, guardrails.

Quick Start

# Prerequisites: jq (JSON processor)
# Windows: winget install jqlang.jq
# Ubuntu/Debian: sudo apt install jq
# macOS: brew install jq

# Add the Claudikins marketplace
/marketplace add elb-pr/claudikins-marketplace

# Install the plugin
/plugin install claudikins-kernel

Restart Claude Code. Then:

# Start your first disciplined session
/claudikins-kernel:outline "Add user authentication to the app"

Meet the Team

These aren't generic "agents". They're your synthetic staff - each with a job and a personality.

The Four Commands

/outline - "Let's figure out what we're building"

Iterative brainstorming until you have a solid plan.

1. Brain-jam - Back and forth with Claude. Pick from options, don't type essays.
2. Research - taxonomy-extremist agents dig through your codebase in parallel.
3. Approaches - 2-3 ways to solve it. Pros, cons, recommendation. You pick.
4. Draft - Plan written section by section. You approve each one.

Output: plan.md with a task table that /execute can parse.

/execute - "Let's build it"

Execute the plan task by task with fresh agents and code review.

1. Batch checkpoint - "Batch 1/3: [task-1, task-2]. Ready?" You decide.
2. Per task - Creates branch, spawns fresh babyclaude, implements, commits.
3. Two-stage review - spec-reviewer checks compliance, code-reviewer checks quality.
4. Merge decision - You choose: merge all, merge some, or keep branches.

Key feature: Each babyclaude gets fresh context. No pollution between tasks.

/verify - "Does it actually work?"

Claude must see the code working. Not trust. Verify.

1. Automated checks - Tests, lint, type check.
2. Output verification - catastrophiser runs your code:
   • Web app? Starts server, takes screenshots.
   • API? Curls endpoints, checks responses.
   • CLI? Runs commands, verifies output.
3. Polish pass - cynic looks for unnecessary complexity. Changes one thing at a time, tests after each.
4. Human checkpoint - Comprehensive report. You decide: ready to ship?

Output: verify-state.json with unlock_ship: true if approved. Plus file hashes so /ship can detect tampering.

/ship - "Send it"

Merge to main with proper docs and PR.

1. Gate check - Won't run unless verify passed AND code hasn't changed.
2. Commit strategy - Squash or preserve? Message drafted, you approve.
3. Documentation - git-perfectionist updates README, CHANGELOG, version. Section by section.
4. PR creation - Draft, approve, create via gh CLI.
5. Merge - Wait for CI if you want. Merge. Cleanup branches.

Output: Code on main. PR merged. Done properly.

The Safety Net

Architecture

Industrial-grade patterns adapted for AI workflows.

Same principles, different scale. Reliability through structure - not speed through parallelism.

Requirements

System

• jq - Used by hook scripts for JSON processing

# Ubuntu/Debian
sudo apt install jq

# macOS
brew install jq

# Windows
winget install jqlang.jq

Recommended Plugins

Optional Plugins

Status

v1.2.0 - Full Release - Regularly Maintained.

View the marketplace | Changelog

License

MIT

We call it Claudikins because "Draconian-AI-Supervisor" was taken.