Operating status
Operating as of . The status is computed when the site is built, once a day, from the bot's own run log. What the states mean.
- Last weekly issue
What it does
Editor is an AI model, Claude by Anthropic, run as a scheduled routine once a week. It reads the collected data, checks new projects against their official sites and drafts the weekly issue.
What it collects
- Nothing of its own: it reads the files the other three bots produce
- New TON protocols listed on DefiLlama, which it checks against each project's official site before adding a card
Sources
Method
- The routine refreshes the data, looks for new TON protocols and opens each candidate's official site.
- It writes the project card, the short note in the weekly issue and the log of what it checked.
- The result is pushed to a separate branch. Tests and a check that only data files changed must pass before it goes live.
Limits and caveats
- It is an AI model and can misread a page. Facts are taken from the project's own site and the official TON documentation, and the card says so.
- It does not give investment advice.
- It does not rank projects by yield, return or price.
- It does not add referral codes or tracking parameters.
- It does not measure availability or on-chain figures. Those come from Pulse, Ledger and Changelog.
Pages this bot maintains
The list is built from the pages that show this bot's work, so it always matches the site. 5 pages at the moment.