# Buzzberg

> Buzzberg turns public market commentary into source-linked research: authors, tickers, directional ideas, sentiment, and author history. Its MCP connection lets compatible assistants query the stored research with the user's account.

When to use Buzzberg: investigate what authors have said about a company or ticker; compare bullish and bearish theses; follow changes in an author's position; review recent ideas, source-linked YouTube research, or a selected ticker portfolio. Choose a bounded research window and limit, then follow the returned pagination and tool-specific instructions. Cite the available original sources and distinguish an author's claim from a verified fact.

Connect to `https://mcp.buzzberg.ai/mcp` using Streamable HTTP and OAuth. Every registered account can connect through browser sign-in and consent. An invitation code or a newly generated personal API key is not required. The required OAuth scope is `mcp`; existing legacy keys remain supported. Public documentation and tool schemas do not grant access to user data or permission to make changes. Use write tools only when the user explicitly requests the change.

Buzzberg is a research tool, not an execution venue or a real-time market-data feed. Price context comes from stored database bars. Do not request provider refreshes, invent missing prices, or equate mention volume with independent author conviction. Account and global rate limits apply: on HTTP 429 honor `Retry-After`, wait, and avoid parallel retry batches. Commercial plans, free-tier quotas and a sandbox are not promised here.

## Connect and discover

- [MCP setup and research examples](https://buzzberg.ai/mcp/index.md): Client setup for Claude, ChatGPT, Codex and Cursor, with the current OAuth flow.
- [Public tool schemas](https://buzzberg.ai/mcp/tools.json): Documentation in the MCP tools/list result shape; names, descriptions, input schemas and available output schemas. Tool execution still requires OAuth.
- [OpenAPI transport description](https://buzzberg.ai/openapi.json): Existing MCP HTTP transport and OAuth discovery; individual research operations use MCP tools/call, not a parallel REST API.
- [OAuth protected resource](https://mcp.buzzberg.ai/.well-known/oauth-protected-resource/mcp): Exact resource identifier, authorization server and scopes_supported.
- [Authorization server metadata](https://mcp.buzzberg.ai/.well-known/oauth-authorization-server): Registration, authorization, token and revocation endpoints and supported methods.
- [Official client documentation](https://github.com/buzzberg-ai/buzzberg-mcp): Maintained setup documentation and compatibility information.

## Understand the product

- [Product overview](https://buzzberg.ai/index.md): What Buzzberg does and links to its public product pages.
- [User guide](https://buzzberg.ai/guide/index.md): Authors, feeds, filters, saved collections and connecting an assistant.
- [About and contact](https://buzzberg.ai/about/index.md): Product background, FAQ and support contact.

## Optional

- [Privacy](https://buzzberg.ai/privacy/index.md): Data handling and account privacy.
- [Terms](https://buzzberg.ai/terms/index.md): Conditions of use.
- [Research disclaimer](https://buzzberg.ai/disclaimer/index.md): Research and educational boundaries.
- [Sitemap](https://buzzberg.ai/sitemap.xml): Indexable site URLs.