Ai

Wraps every Butterbase app's built-in LLM gateway into a single MCP tool: chat completions, embeddings, model listing, config management, and usage reporting. You'd reach for this when your agent needs to call GPT or Claude, generate vectors, or check token spend without leaving the tool context. It's deliberately synchronous, no streaming, because agent tools can't handle partial deltas anyway. The real win is consolidation: one tool covers what would otherwise be five separate API surfaces, and it respects per-app defaults so you're not hardcoding model strings everywhere. If you need streaming or vector search, you'll still need the SDK or the RAG skill.

GitHub stars: 532
last updated: Jun 11, 2026

View on GitHub

Install

$ npx -y skills add butterbase-ai/butterbase-skills --skill ai --agent claude-code

Installs into .claude/skills of the current project.

Files

Butterbase AI Gateway

The AI gateway provides chat, embeddings, model listing, configuration, usage reporting, and asynchronous video jobs. One umbrella tool: manage_ai.

Action	What it does	Returns
`chat`	Synchronous chat completion (no streaming)	OpenAI-shaped `{ choices: [...] }`
`embed`	Vector embeddings for string or string[]	OpenAI-shaped `{ data: [{ embedding: [...] }] }`
`list_models`	Available models with capabilities	`{ models: AiModel[] }`
`get_config`	Current AI config (default model, BYOK key flag, etc.)	`AiConfig`
`update_config`	Set defaults, allowed models, max tokens, BYOK	`AiConfig`
`get_usage`	Token + cost aggregate over a window	usage record
`submit_video`	Submit an ordinary async video job	`{ job_id, status, polling_url }`
`poll_video`	Read an existing video job without submitting again	status, optional stage and authenticated content URLs