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

Quality

Content

77%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

This is a solid, actionable skill that clearly guides experimentation with config variations using specific MCP tools. Its main strengths are the concrete tool usage guidance, the clear clone-with-overrides workflow, and the strong safety guardrails around baseline protection. Its weaknesses are moderate verbosity with some redundancy between sections (Safety vs What NOT to Do vs Step 3 guidance) and some unnecessary explanations of general experimentation principles.

Suggestions

Consolidate the 'Safety: Protect the Baseline' and 'What NOT to Do' sections to eliminate redundancy — the baseline protection guidance appears in both places.

Remove or significantly trim the 'Core Principles' section, as these are general experimentation concepts Claude already understands; the specific enforcement (one-variable-at-a-time via clone overrides) is already well-covered in Step 3.

Dimension	Reasoning	Score
Conciseness	The skill is mostly efficient but includes some unnecessary content. The 'Core Principles' section states things Claude already knows (test one thing at a time, have a hypothesis). The 'What NOT to Do' section partially duplicates the Safety section and Step 3 guidance. The workflow table in Step 2 is helpful but the surrounding prose could be tighter.	2 / 3
Actionability	The skill provides concrete, specific tool names, exact parameter names (sourceVariationKey, key, name, modelConfigKey), precise format examples for modelConfigKey, and clear instructions on which fields to pass vs. omit. The guidance is directly executable with the MCP tools described.	3 / 3
Workflow Clarity	The workflow is clearly sequenced (Steps 1-4) with explicit verification in Step 4, including how to confirm via the clone response or get-ai-config. The safety section provides clear guardrails against destructive operations (never modify baseline). The note on API responses handles a common error recovery scenario.	3 / 3
Progressive Disclosure	The content is well-structured with clear sections and headers, and references related skills at the bottom. However, the document is fairly long for a single file with no bundle files to offload detail into. The modelConfigKey format section, the safety section, and the 'What NOT to Do' section could potentially be organized more compactly or split out. For a standalone skill with no bundle, the organization is decent but not optimal.	2 / 3
	Total	10 / 12 Passed

Description

32%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description conveys a general sense of the skill's purpose—experimenting with configuration variations—but lacks concrete actions, explicit trigger guidance, and distinct terminology. It uses second person ('Helps you'), which violates the voice guidelines, and the absence of a 'Use when...' clause significantly weakens its utility for skill selection among many options.

Suggestions

Add an explicit 'Use when...' clause with trigger terms like 'experiment with configs', 'compare model settings', 'parameter sweep', 'A/B test configurations', or 'try different prompts'.

Replace 'Helps you test' with third-person voice like 'Creates and manages configuration variations to systematically test different models, prompts, and parameters'.

List more specific concrete actions such as 'generate parameter grids', 'compare experiment results', 'track variation performance' to distinguish this from generic configuration or prompt-engineering skills.

Dimension	Reasoning	Score
Specificity	Names the domain (config experimentation) and some actions ('creating and managing variations', 'test different models, prompts, and parameters'), but lacks concrete specific actions like 'create A/B test configs', 'compare results', or 'generate parameter sweeps'.	2 / 3
Completeness	Describes what it does (creating/managing config variations for testing models, prompts, parameters) but has no explicit 'Use when...' clause or trigger guidance. Per rubric guidelines, a missing 'Use when...' clause caps completeness at 2, and the 'what' is also somewhat vague, bringing this closer to 1.	1 / 3
Trigger Term Quality	Includes some relevant keywords like 'configs', 'variations', 'models', 'prompts', 'parameters', and 'experimentation', but misses common user terms like 'A/B test', 'hyperparameter tuning', 'config comparison', 'sweep', or 'ablation'. The terms are somewhat generic.	2 / 3
Distinctiveness Conflict Risk	The concept of 'config variations' and 'systematic experimentation' provides some specificity, but terms like 'models', 'prompts', and 'parameters' are very broad and could overlap with many AI/ML-related skills or general configuration management skills.	2 / 3
	Total	7 / 12 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 11 / 11 Passed

Validation for skill structure

No warnings or errors.

Repository: launchdarkly/ai-tooling
Commit: 24e9c7e

Reviewed: about 7 hours ago

Table of Contents

Discovery Implementation Validation

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.