# Other agents setup

Any agent that supports MCP servers and custom instructions can drive Slideshot. The generic setup is:

1. Connect the agent to the **Slideshot MCP server**.
2. Authorize access with **OAuth** from the agent's interface.
3. Install the raw Slideshot skill files from GitHub in the place your agent expects skills, rules, memories, custom instructions, or project knowledge.

This page is for agents that do not have a dedicated Slideshot setup page yet. If your agent has its own plugin marketplace or MCP installer, use that first.

## 1. Add the Slideshot MCP server

Open your agent's MCP, tools, connectors, or integrations settings and add the **Slideshot MCP server**:

```text
https://api.slideshot.ai/mcp
```

If the agent asks for connection details, use:

- **Name:** `slideshot`
- **Transport:** HTTP or streamable HTTP
- **URL:** `https://api.slideshot.ai/mcp`

The exact labels vary by agent. Some call this MCP, tools, connectors, plugins, integrations, or external servers.

## 2. Authorize access

Connect to the server from the agent's interface. The agent should open an **OAuth authorization** flow in your browser.

Approve access so the agent can call Slideshot tools from the conversation. After approval, the agent should show Slideshot tools such as:

- `create_run`, `get_run`, `list_runs`, `cancel_run`, `submit_run_input`, `list_run_artifacts`
- `list_credentials`, `create_credential`, `update_credential`, `set_default_credential`, `delete_credential`
- `submit_feedback`, `fetch_brand`

## 3. Download the raw Slideshot skill

Download the raw skill files from [`slideshot/skills`](https://github.com/slideshot/skills).

The Slideshot skill lives in this folder:

```text
skills/slideshot/
```

That folder contains the `SKILL` file along with references and supporting instructions the agent needs to use Slideshot correctly.

## 4. Install the skill in your agent

Place the contents of `skills/slideshot/` wherever your agent loads persistent instructions. Depending on the agent, that might be called:

- Skills
- Rules
- Memories
- Custom instructions
- Project knowledge
- System prompt files
- Workspace instructions

If the agent supports **project-level instructions**, install the skill in the project where you want to create demo videos. If it supports **global instructions**, install it globally so Slideshot is available from any workspace.

Restart or reload the agent after adding the files if it does not pick up new skills automatically.

## 5. Test the setup

Ask the agent to record a small workflow:

> Create a Slideshot recording of `https://docs.slideshot.ai` that shows the docs landing page and the getting-started section.

The agent should use the Slideshot MCP tools to create the run, poll until it succeeds, and return the `demo.mp4` artifact.

## Troubleshooting

- **Agent does not support MCP.** Use an agent that supports MCP, or call the [HTTP API](/docs/api) directly from your own integration.
- **MCP server is not visible.** Confirm the URL is exactly `https://api.slideshot.ai/mcp`, then restart or reload the agent.
- **OAuth approval failed.** Re-run the connect or authorize step from the agent's MCP settings.
- **Agent invents endpoint paths or option names.** Reload the skill files from `skills/slideshot/` into the agent's instruction context.
- **Run paused with `awaiting_input`.** Ask the agent to read the prompt and submit the requested value via `submit_run_input`. See [Runs](/docs/api/runs#handling-awaiting-input).
- **Recording stops at a login page.** Save credentials for the target domain. See [Login to any web app](/docs/authentication).

## Related

- [Install skills](/docs/install-skills): how to install the Slideshot skill in any agent that loads instructions.
- [MCP versus CLI](/docs/mcp-vs-cli): when to use the MCP server vs the CLI.
- [API overview](/docs/api/overview): if your agent only speaks HTTP.