Use when building a managed team skills library for a real stack. Map work to shelves, browse before curating, write meaningful `whyHere` notes, and create a starter pack once the first pass is solid.
SkillScore83Maintained
Automatic
◆AI Engineer pack
Last verified 8/16/2026
What it changes
Passed 19 of 23 test cases with the skill, and 12 of 23 without it.
Without skill52%
With skill83%
Checks it passed that the model alone missed
Names the onboarding collection starter-pack.
Maps database tools to the backend shelf area.
Uses add for bundled picks and simple imports.
and 12 more in the full report
8 fixed11 passed either way3 still failing1 broke
Pass-rate lift
+30 pp
Tokens
-5%
Agent turns
0%
Response time
-52%
Model gemini-3.6-flashMeasured 8/16/2026v119 of 23 cases passed with skill
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 8 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 8
Prompt
We are starting a new skill curation project for our team called 'platform-skills'. A developer wants to manually create directories like `skills/` and `docs/` and write `skills.json` by hand using a text editor. What is the standard CLI command sequence to create and enter the workspace?
What was checked, without and with the skill
Checked
Without skill
With skill
Executes npx ai-agent-skills init-library platform-skills to initialize the workspace.
Missed
Passed
Changes directory into platform-skills immediately after initialization.
Passed
Passed
case-07 · gemini-3.6-flash · 8/16/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md130 linessingle file
Curate A Team Library
Goal
Build a managed skills library that another teammate or agent can actually browse, trust, and install.
Do not hand-edit skills.json, README.md, or WORK_AREAS.md when the CLI already has the mutation you need.
First Move
Start with a managed workspace.
npx ai-agent-skills init-library <name>
cd <name>
Ask at most 3 short questions before acting:
what kinds of work the library needs to support
whether the first pass should stay small and opinionated or aim broader
whether the output should stay local or end as a shareable GitHub repo
Shelf System
Use these 5 work areas as the shelf system:
frontend: web UI, browser work, design systems, visual polish
backend: APIs, data, security, infrastructure, runtime systems