> For the complete documentation index, see [llms.txt](https://docs.nimbalyst.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.nimbalyst.com/session-management/commands-and-skills.md).

# / Commands and Skills

Use Claude Code and Codex slash commands, custom commands, and skills in Nimbalyst, and learn how actions, commands, and skills differ.

Slash commands and skills are reusable instructions for your agent: type `/` in the chat and pick one instead of retyping the same request. Nimbalyst supports Claude Code and Codex built-in slash commands, your own custom commands, and skills, all from the same typeahead.

### Actions vs Commands vs Skills

**Actions** are reusable prompt presets defined in a single `nimbalyst-local/ai-actions.md` file and surfaced as a dropdown in the composer. Pick one and the prompt drops into your draft. Best for "I copy-paste this prompt all the time." Typical examples include reviewing changed files, planning an implementation, and drafting release notes. See [AI Actions](/session-management/ai-actions.md).

**Commands** are simple markdown files that provide a prompt template. They live in `.claude/commands/`. When you type `/command-name`, the contents of that file are sent as a prompt to the agent.

**Skills** are richer integrations that can include instructions, workflows, and tool usage patterns. Nimbalyst ships with built-in skills (like `/commit`), plugins can provide skills, and you can create your own. A skill can define how the agent should behave for a specific workflow, for example how to manage your tracker state when working on blog posts.

### Using Slash Commands and Skills

Type `/` in the AI chat to see all available commands and skills. The typeahead shows every skill (user-created, plugin, and extension) so you can quickly find what you need.

Plugin skills are namespaced, and the inserted command matches what the agent SDK routes. For example:

* `/excalidraw:excalidraw`
* `/planning:design`
* `/feedback:bug-report`

Skills and commands written for one agent run in the other: workflow discovery is unified across Claude Code and Codex.

Ready-made skills you can drop into a project are listed on the [Nimbalyst skills page](https://nimbalyst.com/skills/), and the [commands and skills feature page](https://nimbalyst.com/features/commands-and-skills/) covers how they surface in the composer.

### @ Mentions

Type `@` in the AI chat to reference files and folders. The picker opens immediately, with suggestions sorted by recency so the files you most recently opened or edited are at the top. Keep typing to filter, or scroll to find older files. You can mention individual files or entire directories (shown with a folder icon).

Mentioning a file gives the agent direct context about which files you want it to work with, instead of relying on the agent to search for them.

In markdown documents, `@` is also the fastest way to embed supported custom-editor files. If the inserted file link sits by itself on its own line, Nimbalyst upgrades it into a live inline embed in WYSIWYG view. This is useful for mockups, Excalidraw files, data models, CSV files, and other custom-editor content you want to keep inside a spec or plan.

### @@ Session Mentions

Type `@@` in the AI chat to reference another session. The picker shows your sessions in recency order, so the conversation you were just in is one keystroke away.

Mentioning a session pulls its context into your current conversation. Use it when you want the agent to build on prior work, reuse research from an earlier session, or understand decisions made elsewhere, without re-explaining everything.

### Built-in Examples

**`/commit`** creates a git commit:

1. Runs `git status` and `git diff`
2. Reviews recent commits for style
3. Drafts a commit message with type prefix (`feat:`, `fix:`, etc.)
4. Stages and commits

**`/session-cleanup`** tidies your Sessions board:

1. Audits your sessions by phase and tags
2. Suggests phase corrections (for example, moving committed work out of planning)
3. Flags old completed sessions as archive candidates
4. Asks for your approval before changing anything

**`/planning:nimbalyst-coach`** reviews your project and recent sessions, then suggests extensions that match your files, features you have not tried, and instructions worth adding. It changes nothing until you approve.

**`/feedback:bug-report`** and **`/feedback:feature-request`** guide you through filing a bug report or feature request on the public Nimbalyst GitHub repo. See [Feedback, Discord, Support, and Releases](/getting-started/feedback-discord-support-releases.md).

### Create Your Own Commands

Custom commands live in `.claude/commands/` as markdown files. Go to the `.claude` directory in your project, then the `commands` folder, and edit or add your own commands there.

You can also go to Settings (click the gear on the bottom left), then Project Settings, to add commands.

### Create Your Own Skills

Skills are a powerful way to teach the agent how to handle recurring workflows. For example, you can create a skill that integrates with the tracker system, telling the agent to update tracker state whenever you work on certain types of tasks. See the [tracker overview](/task-management/overview.md) for how trackers fit into this.


---

# 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 dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.nimbalyst.com/session-management/commands-and-skills.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

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.