skill-creator
Create or update AgentSkills. Use when designing, structuring, or packaging skills with scripts, references, and assets.
83
Quality
76%
Does it follow best practices?
Impact
92%
2.13xAverage score across 3 eval scenarios
Optimize this skill with Tessl
npx tessl skill review --optimize ./nanobot/skills/skill-creator/SKILL.mdEvaluation results
100%
54%Build a Skill for Automated Invoice Processing
Skill initialization and naming conventions
init_skill.py referenced
0%
100%
Skill folder name format
100%
100%
Skill name is verb-led
0%
100%
Skill folder matches skill name
0%
100%
Frontmatter fields only
0%
100%
Description includes triggers
0%
100%
No when-to-use in body
100%
100%
No auxiliary documentation files
100%
100%
Imperative form in body
0%
100%
No content duplication
100%
100%
Name under 64 characters
100%
100%
Without context: $2.4615 · 7m 58s · 77 turns · 423 in / 29,876 out tokens
With context: $0.6961 · 2m 46s · 28 turns · 32 in / 9,616 out tokens
100%
34%Create a Cloud Deployment Skill Supporting Multiple Providers
Progressive disclosure and multi-variant skill organization
Variant details in references
92%
100%
SKILL.md has selection logic
100%
100%
References linked from SKILL.md
100%
100%
One level deep references
60%
100%
TOC in long reference files
0%
100%
No content duplication
90%
100%
SKILL.md under 500 lines
100%
100%
No auxiliary files in skill
100%
100%
Frontmatter has only name and description
0%
100%
Description includes usage triggers
0%
100%
Without context: $0.9047 · 4m 4s · 28 turns · 1,710 in / 15,065 out tokens
With context: $0.8900 · 3m 36s · 31 turns · 1,926 in / 13,476 out tokens
76%
59%Fix and Ship a Draft Document Editing Skill
Skill packaging and fixing structural issues
package_skill.py referenced
0%
0%
README removed
0%
100%
Extra frontmatter fields removed
0%
100%
Description includes usage triggers
0%
100%
When-to-use removed from body
0%
100%
Content duplication resolved
0%
100%
Script testing documented
0%
0%
Skill folder name preserved correctly
100%
100%
References remain in references/
100%
100%
fix_notes explains rationale
16%
100%
Without context: $0.4268 · 1m 49s · 25 turns · 31 in / 5,880 out tokens
With context: $0.6181 · 2m 17s · 27 turns · 437 in / 7,344 out tokens
- Evaluated
- Agent
- Claude Code
- Model
- Claude Sonnet 4.6
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.