Bumicerts

Most environmental certification is built for big projects. Before a forest restoration project can show its impact, it often has to pay thousands of dollars for auditors and baseline studies. Small Indigenous and local groups, who do much of the world's day-to-day nature care, are priced out before they start. Yet they already document their work: photos on phones, lists in notebooks, reports after every planting day.

Bumicerts are GainForest's attempt to meet communities where they are. "Bumi" means earth in Indonesian and Malay. A Bumicert is a public record of a piece of environmental work, published by the people who did it. It answers three questions:

• Who did the work. The community or organisation, with its own account.
• What they did. A claim that is specific: what, where, when, how much.
• How we know. Evidence: photos, species observations from Taina, AudioMoth recordings, drone maps, tree measurements, reports.

In GainForest's short film about Bumicerts, one of the speakers compares a Bumicert to a book. You start with empty pages and fill them with the story of your work, and you want that book to be "in a very, very, very safe place" and to know "that this book will belong to you."

That is where ATProto comes in. ATProto is the open network that Bluesky runs on, and it can carry any kind of record. Each community has its own account, and its Bumicerts and observations are stored in that account's data store, signed by it. The GainForest app, Ma Earth and other apps can all read the same records. The account can also move to another host, and its signed records move with it, so the community is not locked into GainForest's app or GainForest's servers. One thing to keep in mind: these records are public. That is why the data council from lesson 08 decides what goes in, and why the app lets people hide the exact location of an observation.

We have also learned what Bumicerts do not fix. Between February and November 2025, 46 projects used the first version, then called Ecocerts. They covered agroforestry, mangroves, farming, water, women's leadership and climate education. About two in three gave thin evidence, often a photo or two and a short note. Many of those came from groups with weak connections, no data staff and little support for documenting their work. Projects with a few honest videos and reports did attract more support. Long lists of complex indicators did not reliably attract more.

The money told a harder story. The projects raised $31,724 from 154 contributions, but about 70 percent came from a single funding round. The median contribution was $4.96. A few projects received most of the money and many received almost nothing. A partner in Kalimantan told us donors trusted their claims more because of the certificate, and also that uploading evidence from the interior often failed because the signal was too weak.

So the current version, built on ATProto with the Hypercerts Foundation, aims for fewer, clearer levels of evidence that a small team can actually reach, records that can grow over time instead of one funding moment, and rules for what counts as good evidence that communities help write. A Bumicert is only as good as the trust behind it. The certificate is the cover of the book. The community decides what is written inside.