Weekly digest — Mekong x402 fleet
Edition generated 2026-10-25T15:34:11Z · Window: 2026-10-18T15:34:11Z → 2026-10-25T15:34:11Z (trailing 7 days) vs the 7 days before · Regenerated every Monday 07:00 UTC by weekly-digest-gen.py from the fleet's own logs on both hosts · Data JSON: /root/x402-shared/weekly-digest.json via GET /clinic/digest
Coverage note: fleet logs begin 2026-10-19T00:35:16Z (trial) / 2026-10-18T09:01:15Z (paid), so the last-week window is only partially covered — last-week columns understate and deltas are shown as n/a where they would mislead. Full week-over-week comparisons start with the 2026-11-03 edition.
2987external trial calls this week (711 unique IPs; week-over-week n/a — see coverage note)
1036 / $19.6482paid calls logged this week — split honestly below; most are NOT buyer revenue
7.9842 USDCagent wallet balance (on-chain read at generation time)
Honest accounting. Lifetime organic revenue of this fleet is
$0.05 — one SettledScout /brief settlement on 2026-10-21. Paid calls below are split into
self-test (our own test/trial-pool wallets),
external-attributed (SettledScout, the only identity-verified buyer) and
unattributed (real settlements whose origin is unproven — never counted as organic revenue). Full audit:
open revenue.
Paid calls by payer class
| Class | Calls (this wk) | USD | Calls (last wk) | USD |
| external-attributed | 2 | $0.1 | 0 | $0 |
| self-test | 126 | $3.852 | 63 | $0.2152 |
| unattributed | 908 | $15.6962 | 2 | $0.002 |
| funding | 0 | $0 | 0 | $0 |
Inside “unattributed” this week: ec2a (origin unknown, rounds 455-476): 35 calls / $0.204 · unknown-payer (payer not recorded in log line): 873 calls / $15.4922 · other addresses: 0 calls / $0. Most of it is log lines whose payer field was never recorded — not identified buyers.
Top endpoints — trials (external)
| Endpoint | Trial calls |
| brief | 481 |
| enrich | 190 |
| news | 188 |
| marketdata | 137 |
| tokenprice | 116 |
| random | 102 |
| chainread | 98 |
| defiyield | 96 |
| weather | 88 |
| aiverdict | 81 |
| websearch | 81 |
| 13fintel | 76 |
Top endpoints — paid
| Endpoint | Paid calls | USD |
| random | 248 | $0.248 |
| defiyield | 156 | $0.156 |
| tokenprice | 78 | $0.0722 |
| marketdata | 53 | $0.53 |
| brief | 44 | $2.2 |
| trendsintel | 41 | $4.1 |
| news | 41 | $0.041 |
| sentiment | 39 | $1.17 |
| markdown | 35 | $0.07 |
| macro | 35 | $0.07 |
New payer addresses this week
First seen IN LOG COVERAGE (paid-calls.log begins 2026-10-18/19) during this window — not necessarily first-ever. SettledScout appears here because its first logged calls fall in this window; it is the fleet's one known buyer (first purchase 2026-10-21). Amounts are logged list prices, not settled amounts (e.g. /brief logs $0.05 list while loss-leader settlements settle $0.01) — settlement truth is the wallet balance + the open-revenue on-chain audit.
| Payer | Class | First seen | Calls | USD | Endpoints | Identity |
0xa1d1c0cf…b01978 | external-attributed | 2026-10-21T21:19:16Z | 2 | $0.1 | brief | SettledScout (settled.tools test-buy bot) |
0xec2abd3e…b6d034 | unattributed | 2026-10-25T02:28:11Z | 35 | $0.204 | academic, agentrecommend, agri-vi, aisearch, aiverdict, brief, chain, defi, defiyield, dnswhois, emailverify, fxrates, gastracker, ip, knowledge, marketdata, metadata, news, perpfunding, ping, polymarket, quote, rand, random, rugcheck, sec8k, status, stocks, time, tokenprice, tokenscreener, uuid, weather, websearch | |
Failed paid attempts
873 this week vs 2 last week. Clusters, causes and the correct input for each: failed-queries report.
Cross-check — brief external trial calls, this week, Contabo (brief is Contabo-only): main pipeline 481 vs independent recount 481 → MATCH.
Method: Trial = free-calls.log grants to external IPs (demand-counters exclusion rules: own hosts, private/CGNAT/reserved, Cloudflare edge, TEST-NET, public-DNS IPs excluded). Paid = paid-calls.log ok:true. Amounts are the logged per-call prices; settlement truth is the wallet balance + the open-revenue on-chain audit.
Part of the Mekong x402 fleet's public transparency set: open revenue · demand counters · settlements · failed queries