Tigzig is Built AI-Agent First. 300+ Assets, and the Fastest Way In Is to Ask Your AI.
Published: August 1, 2026
Tigzig is built AI agent first and a fair bit has moved in the past month. You are welcome to browse, but more efficient would be to ask ChatGPT or whichever AI you use ..
If there is anything specific you are looking for then just ask ChatGPT / Claude, Gemini, Meta AI .... something like
- "i want to build excel tool... what does tigzig.com have?"
- "does tigzig.com have any analysis on Donald Trump's trades"
- "what's amar harolikar's view on economy and markets - pull out his posts from tigzig.com"
All 300+ assets on the site, guides, live apps, repos, open APIs and MCP servers, are organized to be machine readable / Agent accessible.
And if you are using an AI Agent / AI Coder (Claude Cowork / ChatGPT Work / Claude Code / Codex), they would be able to pull out the codes, API's and implement, deploy and run it for you ...
Rate limits, and what changed in the traffic
And traffic pattern has changed too. A fair portion of calls now arrive from agents and automated tools, so some callers are running into rate limits they would not have hit earlier.
The rate limits are under active adjustment and you should start to see most of your calls go through.
These would also be documented across all surfaces .. response headers / 429's / machine readable roots and also the human docs. MFPRO API does most of it .. would be extended to other high usage APIs including YFIN & FFN.
If still facing issue, then do consider batching your requests as many API's support multiple tickers / codes ... or for TREMOR and MF API .. you can simply download the whole database .. manually or via API.
API hub: tigzig.com -> API & MCP
Where the traffic is coming from now
And by the way I am getting more and more requests via ChatGPT and Claude.
More people landing on my website via ChatGPT .. after Google that's the number 1 referrer.
And API usage show regular ChatGPT (mostly must be Codex or Work), Claude (must be either CoWork or Codex)... as they send out standard headers and behave in a very predictable and systematic manner .. what i call 'AI agent like'.