Skip to main content
Your agent has four skills from the registry, and three of them weren’t written for this tutorial. The registry has many more. In this part you read one skill’s page, find another, check it before you trust it, and add it to your agent. It all happens in your browser, with nothing to install.
Reading the registry needs no account. To add a skill you need the copy of order-support from Part 1’s step 4.

​
1. Read a skill’s page

Open diagnostic-triage-questions, the skill on your agent’s menu for reports of something broken. Its page shows:
  • What it changes: how many of its test cases passed. What it did to tokens, turns and response time is under Test results. Not every skill has a clear result, and the page says so when it doesn’t. When we wrote this, this skill had passed 13 of 14 cases, and its lift was withheld because that run didn’t record which model graded it.
  • Safety: a code scan, a review of what its instructions try to do, and a content review.
  • SKILL.md (Files, on a skill with several): the instructions your model reads when it opens the skill. Read them: this text goes into your agent’s conversation. When a customer reports something broken, the questions your agent asks come from its list here.
Its page shows no lift, and that doesn’t settle whether it helps you: a score tells you how a skill did on its own tests, not on your agent. Part 5 counts what it changes on yours.

​
2. Search for another

Say you want the agent to handle a complaint it has no rules for: a customer saying Acme broke a service promise. Open the registry and search for sla. Each result shows what the skill says it does, its SkillScore, and whether it passed its safety screening. Read the description, not just the score. Some of these results are about reviewing contract terms, and a search for refund also finds skills that write a returns policy page for a website. They can score well and still be no use to this agent.

​
3. Check it, then add it

Open sla-breach-response. It tells a model how to answer a missed commitment: confirm the miss, concede it, and offer only the remedy the agreement defines. When we wrote this, its page read “Passed 20 of 22 test cases with the skill, and 3 of 22 without it”. That’s a much clearer result than the last skill’s, so read the rest of the page the same way. Check which version it measured: this result is for v1, and adding the skill gets you v2. Check what its test cases ask, too. A strong result on those cases still says nothing about your agent’s questions. Scores give you a shortlist, and your own cases decide. Part 5 shows you how. To add it, go to the skill’s page, open Manage it in DecimalAI, and click Add to an agent. Under Apply to, choose Selected agents, tick order-support, and click Install linked skill. Open your agent’s Prompt & skills page. It now lists five skills. You didn’t change any code or deploy anything: the agent’s skills live on DecimalAI, and its next run reads them. The model decides whether to open the new one, and the kettle’s refund doesn’t call for it, so you’ll see it on the menu rather than in that answer.
A pack is a starting point for one kind of agent: a starter prompt and a set of skills. Skill packs shows how to start from one, or apply one to an agent you already have.decimalai init order-support writes a runnable LangChain file for an agent that already exists in your workspace. --framework openai-agents writes one for that SDK instead, and decimalai init --help lists the rest.decimalai starter support downloads a local Support project with the pack’s prompt and skills copied into it. That copy is frozen: it needs no account and sends no traces, and nothing you change in the dashboard reaches it. The Support starter project walks through it.

Next: 3. Run it in your code

Run this same agent from your own code, with the skill you changed in Part 1.