Anisota storms the ATmosphere

(2025-08-21) We chose a good moment to take another ATmosphere Health Index snapshot yesterday.

anisota.net had said "Hello World" a few hours earlier, and the result was striking, with 76% of the non-Bluesky events (posts, likes, etc.) during our monitoring spotlight attributed to the new app:

While the Anisota launch clearly made quite a splash in the rest of the ATmosphere, it will surprise no one that the vast majority of ATmosphere events during the snapshot (99.92%) were Bluesky events.

How should we measure ATmospheric Health?

We've only been running snapshots over the last few days as we started exploring how to meaningfully measure the ATmosphere's development. We were going to post something a little more polished next month, but Anisota's launch spurred us to reach out to the community for ideas.

The above figures probably provide some of the components of a composite ATmosphere Health Index (AHI):

• overall volume, or total number of events (761,503 measured in yesterday's snapshot) - the more events there are, the bigger the ATmosphere
• app diversity (8, including Bluesky) - the more apps there are, the healthier the ATmosphere
  • note: our snapshot tracked 11, but three had no events during the snapshot period
• non-Bluesky activity (580, or .076%) - more events from non-Bluesky apps indicates a healthier ATmosphere.

As always with composite indexes, however, the devil is in the detail: measuring non-Bluesky activity as a percentage, for example, could result in the overall Index remaining stable despite overall volume dropping, which could not be considered good.

Moreover, the AHI should also probably take into account:

• user-diversity - the more active users these non-Bluesky apps have, the healthier the ATmosphere
• infrastructure-diversity - the more users host their PDSs outside of bsky.social, the more decentralised the network is in practice, making for a more resilient, healthier ATmosphere.

Next steps: your thoughts

We aim to take regular snapshots and publish the raw data and AHI regularly. The goal is to help the community understand where the ATmosphere is going, and build an independent narrative of its evolution based on data, not hearsay.

But we need your input.

The above ideas are still only first thoughts, and many questions remain:

• are we missing any components?
• can we easily collect all of them?
• how often should we take snapshots and publish the results?
• and, most importantly, how do we combine these components into a single, composite index?

The AHI will only be useful if the community thinks it does a good job, so if you have any thoughts we'd love to hear from you. Drop us a line at @atconnect.eu and/or subscribe to the ATConnect newsletter to get, and stay, in touch.