📅 Point-in-time snapshot, sealed 2026-08-02 04:13:21 UTC. A dated checkpoint published during measurement, not a retrospective. For live figures see the economy dashboard.
Read this before quoting any of it
|
2026-08-02 |
| The whole measured economy |
$1,629,783.97 |
| Share going to a single counterparty |
77.7% |
| What remains (agent-to-agent) |
$363,522.98 |
| Of that remainder, wash |
82.71% |
Most of what looks like an economy is one participant. We lead with that rather than bury
it: the concentration is visible in our own Gini (0.995675) and
top-wallet share, so anyone checking would find it in a minute, and a number quoted
without it misleads even when it is arithmetically correct.
The rest of the sealed figures
| Measure |
2026-08-02 |
| Raw on-chain, wash-inclusive |
$1,629,783.97 |
| Validated (raw minus suspect) |
$1,329,101.08 |
| Suspect / wash-attributed |
$300,682.89 |
| Classified consumer commerce |
$1,266,260.99 |
| Real providers (wallets ever paid > $0) |
1,335 |
| Directory listings |
57,429 |
| Distinct payers, all time |
15,151 |
| Wallet coverage |
80.35% |
| Wash-rule version |
3.0 |
Verify this post
Every number above sits inside seal #97, whose chain hash is:
16bd6dbb0d6d5494841db03fe0d937ed853f960e51795de7ac918347c48c76bb
curl -O https://agentalmanac.org/static/verify_integrity.py
python3 verify_integrity.py --witness
That re-derives every hash from the published payloads and cross-checks this site against
our public witness repository,
whose commit dates are recorded by GitHub rather than by us. If we had quietly revised a
past figure, or regenerated the chain, it would fail.
What these numbers do not say
- The concentration figure is arithmetic; the classification is a judgement. We label
the dominant counterparty consumer commerce rather than agent-to-agent settlement, with
the reasoning recorded and versioned by
wash_rule_version (3.0). The headline holds
whether or not you accept that label.
- Raw volume is wash-inclusive. Quote the validated or agent-to-agent figure.
- Do not derive a "dead services" rate from listing counts. Listings counts services we
catalogue, not services we have tested. Only a few hundred have ever been HTTP-probed, so
the rest are untested, not confirmed dead. We do not know the real rate and will not imply
one until we measure it properly.
- Coverage is partial and disclosed at 80.35%, with the
gaps listed on /integrity.
The Agent Almanac measures the agent economy at the settlement layer, on-chain, with the
wash stripped out and the coverage disclosed.