CtrlK
BlogDocsLog inGet started
Tessl Logo

skill-creator

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

Install with Tessl CLI

npx tessl i github:davepoon/buildwithclaude --skill skill-creator
What are skills?

77

Does it follow best practices?

Validation for skill structure

SKILL.md
Review
Evals

Evaluation results

100%

24%

New Skill: Meeting Notes Assistant

Skill initialization with init_skill.py

Criteria
Without context
With context

init_skill.py invoked

100%

100%

Correct CLI syntax

100%

100%

Hyphen-case skill name

100%

100%

Frontmatter name field

100%

100%

Frontmatter description field

100%

100%

Third-person description

33%

100%

Imperative SKILL.md body

91%

100%

Example files removed

0%

100%

Process documented

100%

100%

Skill content relevance

100%

100%

Without context: $0.4887 · 2m 5s · 22 turns · 27 in / 8,034 out tokens

With context: $0.7582 · 2m 50s · 31 turns · 545 in / 9,726 out tokens

97%

19%

Complete the Inventory Database Skill

SKILL.md writing style and resource organization

Criteria
Without context
With context

Imperative body style

75%

100%

Third-person description

0%

100%

Schema in references file

100%

100%

SKILL.md avoids duplication

80%

70%

Script for repeated SQL

100%

100%

SKILL.md references bundled resources

100%

100%

Purpose covered

100%

100%

When to use covered

50%

100%

How to use covered

90%

100%

No placeholder TODOs

100%

100%

Without context: $0.6547 · 2m 4s · 27 turns · 3,228 in / 7,548 out tokens

With context: $0.5459 · 2m · 23 turns · 30 in / 6,876 out tokens

100%

Package the PDF Utilities Skill for Distribution

Skill packaging with package_skill.py

Criteria
Without context
With context

package_skill.py invoked

100%

100%

Correct CLI syntax

100%

100%

Validation errors fixed

100%

100%

Zip file produced

100%

100%

Fix documented

100%

100%

Script run from scripts dir or via python

100%

100%

No manual zip fallback

100%

100%

Without context: $0.3045 · 1m 8s · 17 turns · 24 in / 4,030 out tokens

With context: $0.5951 · 2m 1s · 25 turns · 29 in / 6,727 out tokens

Evaluated
Agent
Claude Code

Table of Contents

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.