ce-optimize

Automate end-to-end optimization experiments across code and config variants.

2|Updated Apr 30, 2026

Search Tags:#worktree #baseline #optimize #measurement #reproducibility #experimentation #judge-evaluation

Ready to run

npx skills add https://github.com/thisisryanswift/zellij-agent-tools --skill ce-optimize-thisisryanswift

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: ce-optimize
Source: https://github.com/thisisryanswift/zellij-agent-tools/tree/main/.opencode/skills/ce-optimize
Command: npx skills add https://github.com/thisisryanswift/zellij-agent-tools --skill ce-optimize-thisisryanswift

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires python3, and includes references (resource) components.

What problem does it solve?

Iterative optimization loops that coordinate reusable measurement harnesses, reproducible baselines, and a backlog of hypotheses to discover the best code/config variants.

Core Features & Use Cases

Coordinated setup of baselines and experiments with crash-safe logging and persistent experiment logs
Batch evaluation with hard metrics or LLM-as-judge scores and automatic judge evaluation
Reproducible workspace management via git worktrees and scratch spaces to isolate experiments

Quick Start

Provide a complete optimization spec and launch the workflow to start with a baseline and first batch.

Frequently Asked Questions about ce-optimize

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

FAQPage Schema

How do I coordinate automated optimization experiments across code variants?▼

Automated optimization experiments are coordinated using a repeatable measurement harness that evaluates code and config variants against a baseline. It requires a valid optimization spec and manages parallel evaluation with crash-safe logging to preserve results.

How does git worktree workspace management work for reproducible experimentation?▼

Git worktree workspace management isolates reproducible experiments by creating separate scratch spaces for each variant. This ensures measurement harnesses run consistently without cross-contamination between parallel optimization evaluations.

Can I use LLM-as-judge scores for batch evaluation of optimization hypotheses?▼

Batch evaluation supports LLM-as-judge scores alongside hard metrics. The system includes automatic judge evaluation, requiring optional judge configuration within your optimization spec to score and compare code variants.

What do I need to set up before launching an optimization experiment workflow?▼

You need Python3 installed, a complete optimization spec including name, metric, and measurement definitions, plus a defined mutable and immutable scope. Optional judge configuration and approved dependencies can also be specified.

Why do I need a defined mutable and immutable scope for optimization experiments?▼

Defining mutable and immutable scopes isolates which code or config variants the optimization experiments can alter. This constraint ensures the baseline measurement harness remains reproducible across parallel evaluations without unintended side effects.

What happens to my experiment logs if the optimization evaluation process crashes?▼

Experiment logs are preserved through crash-safe logging that persists results directly to disk. This ensures your baseline data and completed batch evaluations survive interruptions during parallel optimization workflows.