configs-variations

Experiment with configs by creating and managing variations. Helps you test different models, prompts, and parameters to find what works best through systematic experimentation.

2.75x

Quality

55%

Does it follow best practices?

Impact

99%

2.75x

Average score across 3 eval scenarios

Securityby

Passed

No known issues

Optimize this skill with Tessl

npx tessl skill review --optimize ./skills/agentcontrol/configs-variations/SKILL.md

Evaluation results

98%

65%

Cost Reduction Experiment for Customer Support AI

Clone variation with model override only

Criteria

Without context

With context

Uses clone tool

100%

modelConfigKey format

100%

Only model fields overridden

100%

sourceVariationKey provided

100%

New variation key and name

100%

Inheritance explained

50%

100%

Single variable changed

70%

100%

Hypothesis stated

100%

Result reporting

20%

60%

No retry on success

100%

80%

Model Migration for Email Draft Assistant

Protect baseline on replace request

Criteria

Without context

With context

Creates new variation

100%

Uses clone tool

100%

Does not update baseline

100%

Does not delete baseline

100%

Explains rationale to user

100%

Correct modelConfigKey format

100%

New variation key and name

100%

sourceVariationKey provided (clone path)

100%

Only model fields overridden

100%

Final state described

100%

45%

Multi-Hypothesis Experiment for Code Review Bot

One-variable experiment design

Criteria

Without context

With context

Separate variations per hypothesis

100%

Uses clone tool

100%

Model variation: only model fields

50%

100%

Parameter variation: only parameters

50%

100%

Instructions variation: only instructions

50%

100%

modelConfigKey format

100%

Variation keys and names

100%

sourceVariationKey for each clone

100%

Baseline not modified

100%

Experiment rationale explained

100%

Repository: launchdarkly/ai-tooling
Commit: 24e9c7e

Evaluated: about 7 hours ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Cost Reduction Experiment for Customer Support AI Model Migration for Email Draft Assistant Multi-Hypothesis Experiment for Code Review Bot

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.