Skip to content

Machine access

Chain of Thought for AI agents

AI agents can search Chain of Thought episodes and transcript passages through the public MCP server at https://mcp.chainofthought.show/mcp. Each result gives a chainofthought.show link. Cite the episode page, or the linked moment when referring to a passage.

MCP server

The remote endpoint uses Streamable HTTP. It is stateless and read only. You do not need an account, login, or API key. The server code is available under MIT. Transcript text and episode metadata are licensed CC BY 4.0: credit Chain of Thought, link the episode page and the license, and say if you changed the text. Episode audio and video are not licensed: all rights reserved. The transcripts are also available as a dataset.

Available MCP tools
ToolWhat it returns
search_transcriptsFull-text search over transcript passages, up to 20 results with a link to each moment.
search_episodesEpisodes matching a topic, guest, company, title, or chapter.
get_episodeEpisode metadata, guests, chapters, and links. It does not return the transcript.
get_transcript_excerptUp to 16 consecutive transcript segments around a point.

Connect

Claude Code

claude mcp add --transport http chain-of-thought https://mcp.chainofthought.show/mcp

Claude on claude.ai or Desktop

Open Settings > Connectors > Add custom connector, then paste https://mcp.chainofthought.show/mcp.

ChatGPT

Where your plan supports custom connectors, add the remote MCP URL in developer mode connectors: https://mcp.chainofthought.show/mcp.

Cursor

Add this entry to mcp.json:

{
  "mcpServers": {
    "chain-of-thought": {
      "url": "https://mcp.chainofthought.show/mcp"
    }
  }
}

Other machine-readable endpoints

These public files come from the same episode archive as the HTML pages. Use the manifest to find current episode JSON URLs.

  • /llms.txt: an index of citable episode and reference pages.
  • /llms-full.txt: the episode map with takeaways and FAQ text.
  • /corpus.json: a manifest with episode metadata, counts, usage guidance, and each episode JSON URL.
  • /corpus.jsonl: newline-delimited transcript segments, each with a citation URL.
  • Open dataset (Hugging Face, CC BY 4.0): transcript text and episode metadata as Parquet, without audio or video. It is a snapshot refreshed periodically, so the newest episodes can be missing from it.
  • /podcast/{slug}.json: one episode's metadata, chapters, and transcript. Follow the json field in the manifest for the current path.

How to cite

Cite the episode page URL supplied with a result. For a specific statement, use its moment link with the fragment anchor. The canonicalUrl field in corpus data points to the same HTML citation surface. Transcripts are machine-generated and can contain errors, so verify quotations against the recording before publishing them.

Limits and freshness

The MCP server limits each IP address to 10 requests per 10 seconds. Results can be cached for up to an hour. A new episode is usually searchable within about an hour after it appears on the site.

Privacy

The MCP server is operated by Fernwillow Holdings LLC, the company that runs Chain of Thought. This section covers the server at mcp.chainofthought.show. It has no accounts, sets no cookies, and keeps no query history in its database.

  • What it receives. Each request carries the tool arguments, such as search terms or an episode number, plus the network details every web request has, such as your IP address. When you use the server through an AI app such as ChatGPT or Claude, the app sends the tool arguments for you. The server asks for no account details, but anything you put in a search reaches it, so leave out personal or sensitive information. The connecting IP address may be the app's rather than yours. The app's own privacy policy covers what it keeps.
  • Why. To answer the request, to enforce the limit of 10 requests per 10 seconds per IP address, and to find and fix errors.
  • Where it goes and how long it stays. Cloudflare hosts the server. Results can be cached at Cloudflare's edge for up to an hour, keyed by the normalized request, which includes the search terms but not your IP address. Cloudflare's request logs for the server record request metadata, and the server adds a line per request with the tool name, whether it succeeded, the result count and the response time, never the search terms. These logs are kept for at most 7 days. Cloudflare's security and traffic analytics, which apply the rate limit, record IP addresses and keep them for up to 31 days. Cloudflare processes this data only to host, cache and protect the server.
  • What it isn't used for. Fernwillow Holdings LLC doesn't sell request data or use it for model training, profiling or advertising.
  • Your choices. You can stop sending requests at any time by disconnecting the server in your AI app. To ask about your data, request deletion, or ask about this policy, email podcast@conorbronsdon.com. We'll tell you what can be found and delete it where we can; cached results and logs also expire on their own on the schedule above.

The server isn't directed at children. Last updated October 14, 2026.

Terms of use

The MCP server and the data endpoints on this page are free to use. Fernwillow Holdings LLC provides them as they are, without warranties, and may change, rate-limit or stop them at any time. Transcripts are machine-generated and can contain errors, so check a quote against the episode before relying on it.

Transcript text and episode metadata are licensed CC BY 4.0: credit Chain of Thought, link the episode page and the license, and say if you changed the text. Episode audio and video are not licensed: all rights reserved. Don't try to get around the rate limit or disrupt the service; for bulk access, use the open dataset or /corpus.jsonl instead.

To the extent the law allows, Fernwillow Holdings LLC is not liable for losses that come from using the server or its data. Questions go to podcast@conorbronsdon.com.