> For the complete documentation index, see [llms.txt](https://docs.uptiq.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.uptiq.ai/ai-workbench-qore/platform-resources/skill-library.md).
# Skill Library
The skills the platform provides, and the skills you configure for an agent.
A **skill** is a single action an agent can take. The skills attached to an agent are its complete set of actions: the planner chooses from exactly that set, whatever the instructions say.
There are three kinds of skill. Skills are set per agent, so turning one on or adding one changes only that agent.
| Kind | Where you find it | What you do with it |
| ------------------------ | --------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------- |
| **Built-in skills** | The **Pre-attached Skills** tab of [Add Skill](/ai-workbench-qore/console/agent-builder/build/adding-a-skill-to-the-agent.md), and the agent's **Skills** panel | Enable the ones the agent needs |
| **Skills you configure** | The **Skill Library** tab of [Add Skill](/ai-workbench-qore/console/agent-builder/build/adding-a-skill-to-the-agent.md) | Add the skill, then configure it for your system or logic |
| **Custom skills** | The **Custom Skills** tab of Add Skill | Upload a folder with a `SKILL.md` file. See [Custom skills](/ai-workbench-qore/platform-resources/skill-library/custom-skills.md) |
## Built-in skills
Every agent has the same built-in skills. The **Skills** panel shows how many in each group are active, for example **8/15 active**. Active skills are available to the planner. Disabled skills are not.
In the **Skills** panel:
* Each group lists its active skills first. The rest appear under a heading with the disabled count, for example **7 Disabled**.
* Use **Search Skills** to find a skill by name.
* Use **Select** to choose several skills and turn them on or off together.
### Enable a built-in skill
1. On the Builder Canvas, select **+ Add Skill**.
2. Open the **Pre-attached Skills** tab. It lists **Active Skills** first, then **Skills you can add**.
3. Find the skill under **Skills you can add**, and select **Enable**.
The skill moves to **Active Skills**, and the planner can use it. Each active skill has an **Actions** menu.
| Group | Description | Open |
|---|
| General skills | Reasoning, web search and crawling, knowledge search, code execution, files, and messages. | General skills |
| Document skills | Extract, classify, summarize, analyze, display, and generate documents. | Document skills |
{% hint style="info" %}
**Turn off what the agent doesn't need.** Each active skill is one more option the planner has to choose between. Keep active only the skills the agent's job requires.
{% endhint %}
## Skills you configure
Add these from the **Skill Library** tab when the agent needs to reach your own systems or run your own logic. The tab lists these skills, all by Uptiq:
| Skill | What it does |
| ---------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------- |
| **MCP Server** | Connects to a third-party MCP server, so the agent can use its tools and services |
| [**API Call**](/ai-workbench-qore/platform-resources/skill-library/api-call.md) | Calls an external service over HTTP to fetch live data |
| [**API Call Advanced**](/ai-workbench-qore/platform-resources/skill-library/api-call.md#api-call-advanced) | Calls an external service over HTTP, with dynamic variables in headers, query parameters, path parameters, and the request body |
| **JavaScript Code** | Runs JavaScript code in the agent |
| **Python Code** | Runs Python code in the agent |
| **Structured Data Query** | Answers natural-language questions from a database, by converting them to SQL and running it |
| **Web Crawler** | Visits and scrapes websites to collect structured or unstructured data for the agent |
| **Document Extraction** | Extracts key information, tables, and text from PDFs, Word files, or images. Enabled by default |
| **Document Summarizer** | Summarizes a document, extracting its key information into a short overview. Enabled by default |
Give each skill a clear name and description. The planner reads them to decide when to use the skill.
### What each skill asks for
{% tabs %}
{% tab title="API Call" %}
A URL and method, plus authorization, headers, query parameters, and input and output schemas. **API Call Advanced** adds a request body. See [API Call](/ai-workbench-qore/platform-resources/skill-library/api-call.md).
{% endtab %}
{% tab title="MCP Server" %}
Selecting **MCP Server** opens **Create MCP Connection**, which starts with a reminder: the connector isn't verified by Uptiq, so confirm that you trust it before you use it.
| Field | What to enter |
| ------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Connection Name** | Required. A name for the connection |
| **Remote MCP Server URL** | Required. The server's URL |
| **Advanced Settings** | Optional. The **Redirect URL** to register as an allowed redirect URI in your MCP server's OAuth app; **OAuth Client ID** and **OAuth Client Secret**; **Authenticate M2M (OAuth client credentials)**, with a **Token URL** and optional **Scope**; and **Headers**, as key and value pairs |
To fill the form from a file instead, select **Import Config File**. Select **Next** to continue. If the server needs a sign-in, you authorize it in a pop-up; then you choose which of its tools the agent may use, and select **Create Connection**. See [MCP Tools](/ai-workbench-qore/console/agent-builder/build/mcp-tools.md#connecting-an-mcp-server).
{% endtab %}
{% tab title="Web Crawler" %}
**Web Crawler** is a ready-made MCP connection to Firecrawl. Selecting it opens **Create MCP Connection** with **Connection Name** set to **Web Crawler** and **Remote MCP Server URL** set to `https://mcp.firecrawl.dev/{{api_key}}/v2/mcp`. Replace `{{api_key}}` with your Firecrawl API key, either the key itself or a reference to a secret, such as `$secret.FIRECRAWL_API_KEY`. Then select **Next**. You provide the key: Firecrawl offers free and paid keys.
{% endtab %}
{% tab title="JavaScript Code and Python Code" %}
Both open a full-screen code editor.
| Field | What to enter |
| -------------------------------------- | ------------------------------------------------------------------------ |
| **Name** and **Description** | What the planner calls the skill, and when to use it |
| **Timeout** | How long the code may run, in milliseconds |
| **Code** | The code. The editor starts with a `main` function that returns a result |
| **Input Schema** and **Output Schema** | The skill's input and output, in **Basic** or **Advanced** view |
Switch between **Editor**, where you write the code and select **Save**, and **Try Out**, where you enter test input and select **Execute** to see the output.
{% endtab %}
{% tab title="Document Extraction and Document Summarizer" %}
Both are enabled by default, so there is nothing to configure. They are the same skills as the built-in **Extract Document Data** and **Summarize Document**. See [Document skills](/ai-workbench-qore/platform-resources/skill-library/document-skills.md).
{% endtab %}
{% endtabs %}
The Add Skill dialog also has tabs for **Flows**, **Agents**, and the **MCP Marketplace**. See [Adding a skill to the agent](/ai-workbench-qore/console/agent-builder/build/adding-a-skill-to-the-agent.md).
## Related pages
* [Skills & Tools](/ai-workbench-qore/core-concepts/skills-and-tools.md): what skills are and how the planner chooses between them.
* [Skills & Capabilities](/ai-workbench-qore/console/agent-builder/build/skills-and-capabilities.md): the full list of skills on one agent.
* [PII Masking](/ai-workbench-qore/console/agent-builder/deploy/guardrails/pii-masking.md): masking personal data, which is an agent guardrail rather than a skill.
***
{% columns %}
{% column width="83.33333333333334%" %}
Maintained by Abhishek Paul
AI-assisted, human-approved
{% endcolumn %}
{% column width="16.666666666666664%" %}
{% endcolumn %}
{% endcolumns %}
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.
Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:
```
GET https://docs.uptiq.ai/ai-workbench-qore/platform-resources/skill-library.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
