The AI execution system that stops your coding agent from going rogue. Rules, skills & workflows for Antigravity, Claude Code, and Cursor — install in one command.
<div align="center">
# KELAR
**Kept Efficient, Logical, Atomic, Resilient**
*The AI execution system for developers who are done babysitting.*
*Sistem eksekusi AI untuk developer yang sudah bosan ngatur-ngatur AI-nya.*
[](https://npmjs.com/package/kelar-cli)
[](./LICENSE)
[](.)
> ```npx kelar-cli@latest init```
</div>
---
## The Problem
AI coding tools fail in predictable ways. You ask for a feature — it touches files you didn't mention, hardcodes values, ignores your patterns. You ask it to fix a bug — it wraps the symptom in a try-catch. You hit a context limit — it greets you like a stranger.
KELAR fixes that. It's a layer of **rules, skills, agents, and workflows** that sits on top of your AI agent and makes it behave like a senior developer.
---
## How It Works
KELAR has four layers. Each one does a different job.
```mermaid
graph TD
A[🔴 Rules<br/>Always-on guardrails] --> E[Your Codebase]
B[🟡 Skills<br/>Context-triggered protocols] --> E
C[🟢 Workflows<br/>Structured command flows] --> E
D[🔵 Agents<br/>Specialized sub-AI spawned per task] --> E
style A fill:#1a1a1a,stroke:#ef4444,color:#fff
style B fill:#1a1a1a,stroke:#eab308,color:#fff
style C fill:#1a1a1a,stroke:#22c55e,color:#fff
style D fill:#1a1a1a,stroke:#3b82f6,color:#fff
style E fill:#0f0f0f,stroke:#333,color:#888
```
```mermaid
graph LR
subgraph state [".kelar/state/ — living memory"]
S1[STATE.md]
S2[TASKS.md]
S3[PATTERNS.md]
S4[HANDOFF.md]
end
subgraph memory [".kelar/memory/ — knowledge base"]
M1[domain/]
M2[technical/]
M3[solutions/]
end
subgraph tools ["kelar-tools.cjs — CLI utility"]
T1[state ·Agent that generates comprehensive documentation, API references, architecture diagrams, and developer onboarding guides from existing code.
Agent configuration for systematic bug investigation that traces issues from error logs through the codebase to root cause with suggested fixes.
Agent for integrating third-party APIs including SDK setup, type generation, error handling, retry logic, and rate limit management.
Cursor's built-in autonomous coding agent that can make multi-file edits, run terminal commands, search the codebase, and iteratively build features with minimal human intervention.
Cloud-based autonomous coding agent that runs in the background on remote sandboxed environments, handling complex multi-step tasks while you continue working.
Cursor's multi-file editing agent within Composer mode that can create, edit, and delete files across your entire project in a single conversation.