> For the complete documentation index, see [llms.txt](https://docs.stepsecurity.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.stepsecurity.io/github/ai-agents/agent-plugins.md).

# Agent Plugins

The **Agent Plugins** page provides an inventory of the AI agent plugins published from the repositories in your GitHub organization. It shows each plugin StepSecurity detects from its plugin manifest, the AI agents it is published for, the repository and folder it is published from, its declared version, the repositories whose settings enable it, and the components it supplies.

To open it, select **AI Agents** > **Agent Plugins** in the **GitHub** section of the left navigation.

{% hint style="info" %}
This page covers plugins published from your GitHub repositories. For plugins installed on developer machines, see [Agent Plugins](https://docs.stepsecurity.io/developer-machines/ide-and-ai-agents/agent-plugins.md) in Dev Machine Guard.
{% endhint %}

A plugin is a bundle. A single install can add skills, commands, hooks, MCP (Model Context Protocol) servers, and other components to an AI coding agent at once, and those components run with the privileges of the developer running the agent. A repository that publishes a plugin is a distribution point: developers install the plugin from it onto their own machines. A change merged into a published plugin's folder is what those developers receive the next time they install or update the plugin, so repositories that publish plugins are part of your software supply chain.

The page header shows the number of plugins and the number of repositories they are published from.

<figure><img src="https://754495266-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FQJRZY4cfEeY3I7DXTOCp%2Fuploads%2FK05NEbzyQZCRd3UIjoIQ%2FScreenshot%202026-10-19%20at%2015.01.04.png?alt=media&amp;token=09fda33e-b38f-45d8-8092-1be7871b65f2" alt=""><figcaption></figcaption></figure>

### How Plugins Are Detected

StepSecurity identifies a plugin by its manifest file. Each supported agent has its own manifest:

| Agent       | Manifest                     |
| ----------- | ---------------------------- |
| Claude Code | `.claude-plugin/plugin.json` |
| Codex       | `.codex-plugin/plugin.json`  |

The folder that contains the manifest directory is the **plugin folder**. A plugin folder can be the repository root or a subfolder, for example `plugins/devtools/`, so one repository can publish several plugins. A plugin folder that holds manifests for more than one agent appears once in the inventory, with every agent it is published for.

Each plugin records the branch and commit StepSecurity read it from, and when it was last scanned.

#### Filtering and Search

**Agent chips** scope the list to one AI agent: `Claude Code`, `Codex`, or `GitHub Copilot`. `All` shows every plugin. Each chip shows how many plugins are published for that agent. A plugin published for more than one agent counts under each of its agents, so the agent counts can add up to more than the `All` count.

Below the chips, you can:

* Search plugins or their components by name
* Narrow the table to one repository with the **All repositories** dropdown

#### Plugin Table

Each row is one plugin folder in one repository. You can sort the table by **Plugin**, **Used by**, and **Components**.

| Column                | Description                                                                                                                                      |
| --------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------ |
| **Plugin**            | The plugin name, with the author declared in its manifest beneath                                                                                |
| **Agent**             | The AI agents the plugin is published for, one badge for each manifest found in the plugin folder                                                |
| **Repository · path** | The repository the plugin is published from, with the plugin folder beneath. A plugin at the top level of the repository shows `repository root` |
| **Version**           | The version declared in the plugin's manifest, for example `1.0.0`. A dash means no version was read from the manifest                           |
| **Used by**           | The repositories in your organization whose settings enable the plugin. A dash means no repository enables it                                    |
| **Components**        | What the plugin supplies, counted by kind, for example `5 skills`                                                                                |

### Plugin Details

Select a plugin to open its details panel. The header shows the plugin name, a badge for each agent it is published for, and its version, followed by the author and repository, for example `Agent Plugin · Example Org · published from engineering-plugins`.

<figure><img src="https://754495266-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FQJRZY4cfEeY3I7DXTOCp%2Fuploads%2FYns8iKrdnWADhAJRSa5c%2FScreenshot%202026-10-19%20at%2015.01.11.png?alt=media&amp;token=5df470d6-05bb-45e0-9fff-254f1b0fdfa0" alt=""><figcaption></figcaption></figure>

Cards at the top summarize the plugin:

| Card             | What it shows                                    |
| ---------------- | ------------------------------------------------ |
| **Components**   | The number of components the plugin supplies     |
| **Agents**       | The number of agents the plugin is published for |
| **Last scanned** | When StepSecurity last scanned the plugin        |

#### **Plugin Information**

The plugin's **Description**, **Name**, **Version**, **Author**, **License**, **Homepage**, and **Declared repository**, as declared in its manifest.

{% hint style="warning" %}
Author, homepage, and declared repository are self-declared manifest metadata. A manifest can name any author or repository. Use **Source** to see the repository and commit the plugin is actually published from.
{% endhint %}

#### **Source**

Where the plugin is published from. Developers install the plugin from this location onto their machines. Repositories that enable the plugin for their own agent sessions appear separately, under **Used by**.

* **Repository**: the repository that publishes the plugin, linked to GitHub
* **Plugin folder**: the plugin's folder in the repository, linked to its location. A plugin at the top level of the repository shows `repository root`
* **Manifests**: each manifest found in the plugin folder, labeled with its agent and linked to the file in the repository, for example `.claude-plugin/plugin.json` for Claude Code and `.codex-plugin/plugin.json` for Codex
* **Branch**: the branch the plugin was read from
* **Commit scanned**: the commit StepSecurity scanned, with a copy button

#### **Used by**

The repositories in your organization whose settings enable the plugin. An entry here means the plugin is configured for everyone who works in that repository. Whether a developer has installed the plugin on their machine is reported by Dev Machine Guard on [Agent Plugins](https://claude.ai/developer-machines/ide-and-ai-agents/agent-plugins.md).

When no repository in the organization enables the plugin, the section says so.

#### **Components**

Every component the plugin supplies, grouped by kind, such as **Skills**. Each component is listed by name and linked to its location in the repository. Hooks are listed by the event they run on.

{% hint style="info" %}
Review a plugin's components before your organization enables it or recommends it to developers. Hooks run automatically on agent events, MCP servers give the agent access to external systems, and bundled scripts run when the agent invokes them. Whatever runs, runs with the privileges of the developer running the agent.
{% endhint %}


---

# 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.stepsecurity.io/github/ai-agents/agent-plugins.md?ask=<question>&goal=<user_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.
