{"name":"Agentic Finance Graph — changepoints","description":"Dated changes to coverage, method, the asset set, the label map and detector rules, with the metric keys each one moved. A step in a series at one of these dates is our change, not the market's.","rule":"Entries are never edited or removed; a wrong entry is corrected by a later one.","license":"https://creativecommons.org/licenses/by/4.0/","attribution":"Agentic Finance Graph — https://agenticfinancegraph.com","count":10,"changepoints":[{"id":"cp-2026-09-27-sweep","date":"2026-09-27","kind":"coverage","comparable_from":"2026-10-03","metric_keys":["actors_l7_plus","actors_l8_plus","actors_l9","observed_pay_events","observed_pay_usd_total","observed_pay_counterparties","observed_pay_usd_median","observed_pay_top_actor_share","pass_through_pay_usd_excluded","pre_registration_pay_usd_excluded","bound_wallet_usdc_median","bound_wallet_usdc_total","bound_wallet_funded_rate","counterparty_top_payer_share","counterparty_contract_share"],"definition_ids":["rank.l7plus.v1","rank.l8plus.v1","rank.l8plus.v2","rank.l9.v1","rank.l9.v2","pay.usd.total.v1"],"summary":"Every bound wallet is swept every three hours for USDC, USDT, EURC and DAI transfers (before: a smaller watched set, USDC only), and the per-wallet scan reads up to 10,000 transfers instead of the newest 50. The counts rose because more of the existing history became visible, not because agents began paying that day: L7+ went 519 → 2,387 on 27 Sep and 7,681 by 3 Oct, when the backfill settled. Compare these series from 3 Oct 2026.","reference":"https://agenticfinancegraph.com/changelog-september-2026"},{"id":"cp-2026-10-11-detectors","date":"2026-10-11","kind":"detector_rule","comparable_from":null,"metric_keys":["detections_logged","detections_7d"],"definition_ids":["detector.retry_storm.v2","detector.mint_burst.v2","detection.logged.v1"],"summary":"retry_storm and mint_burst moved from hour buckets to runs (rule v2), and the layering_fan detector was added. 35 storm runs and 71 burst runs the old rules split or missed were added to the log at about 14:40 UTC, so detections_logged stepped 610 → 716.","reference":"https://agenticfinancegraph.com/changelog-october-2026"},{"id":"cp-2026-10-12-labels","date":"2026-10-12","at":"2026-10-12T09:16:04Z","kind":"label_map","comparable_from":null,"metric_keys":["observed_work_escrow_pay_events","observed_work_escrow_pay_usd","observed_yield_vault_pay_events","observed_yield_vault_pay_usd","observed_lending_pay_events","observed_lending_pay_usd","observed_bridge_pay_events","observed_bridge_pay_usd","observed_swap_pay_events","observed_swap_pay_usd","counterparty_labelled_share"],"definition_ids":["cp.cat.swap.usd.v1","cp.cat.lending.usd.v1","cp.cat.bridge.usd.v1","cp.cat.work_escrow.usd.v1","cp.cat.yield_vault.usd.v1","cp.labelled.share.v1"],"summary":"The counterparty label map was updated: newly named addresses moved existing payments into categories. Swaps stepped 662,460 → 927,257 USD and the labelled share 0.505 → 0.525 in the next run. No new payments caused the step.","reference":"https://agenticfinancegraph.com/changelog-october-2026"},{"id":"cp-2026-10-09-dq-checks","date":"2026-10-09","kind":"method","comparable_from":null,"metric_keys":["dq_checks_total","dq_warn_failing","dq_block_failing","dq_health_score"],"definition_ids":["dq.checks.v1","dq.health.v1"],"summary":"Three data checks were added: 37 → 40 checks. A step in the health score on this date comes from the new checks, not from a change in the data.","reference":"https://agenticfinancegraph.com/data-health"},{"id":"cp-2026-10-12-dq-check","date":"2026-10-12","at":"2026-10-12T12:00:00Z","kind":"method","comparable_from":null,"metric_keys":["dq_checks_total","dq_warn_failing","dq_block_failing","dq_health_score"],"definition_ids":["dq.checks.v1","dq.health.v1"],"summary":"One data check was added: 40 → 41 checks.","reference":"https://agenticfinancegraph.com/data-health"},{"id":"cp-2026-10-12-chain-lag","date":"2026-10-12","at":"2026-10-12T16:35:00Z","kind":"method","comparable_from":null,"metric_keys":["dq_checks_total","dq_warn_failing","dq_health_score"],"definition_ids":["dq.checks.v1","dq.health.v1"],"summary":"A data check for lagging chain watchers (chain_lag, warn) was added: 41 → 42 checks. The health score fell below 1.00 because a real lag (Arbitrum) is now counted, not because data got worse.","reference":"https://agenticfinancegraph.com/changelog-october-2026"},{"id":"cp-2026-09-27-sweep-scope","date":"2026-09-27","kind":"coverage","comparable_from":"2026-10-03","corrects":"cp-2026-09-27-sweep","registered":"2026-10-12","metric_keys":["counterparties_known","counterparties_with_3plus_payers","counterparties_labelled","counterparty_labelled_events","counterparty_labelled_share","observed_bridge_pay_events","observed_bridge_pay_usd","observed_cross_chain_pay_events","observed_cctp_exit_events","observed_cctp_exit_usd","observed_lending_pay_events","observed_lending_pay_usd","observed_work_escrow_pay_events","observed_work_escrow_pay_usd","observed_yield_vault_pay_events","observed_yield_vault_pay_usd","observed_swap_pay_events","observed_swap_pay_usd","observed_x402_pay_events","observed_x402_pay_usd","observed_x402_share_of_payments","reputation_rated_l7_actors","idle_funded_actors","idle_funded_usd"],"definition_ids":["pay.usd.total.v1"],"summary":"Scope correction to cp-2026-09-27-sweep: the series computed from the same counted payments (counterparties, payment categories, the x402 rail share, rated L7 actors, idle funded actors) stepped with it between 27 Sep and 3 Oct, while the stored payment rows grew from 15,393 to 277,120. These steps are the backfill, not new activity. Compare these series from 3 Oct 2026.","reference":"https://agenticfinancegraph.com/changelog-september-2026"},{"id":"cp-2026-10-12-unexplained-step","date":"2026-10-12","at":"2026-10-12T21:46:00Z","kind":"method","comparable_from":null,"metric_keys":["dq_checks_total","dq_checks_passing","dq_warn_failing","dq_health_score"],"definition_ids":["dq.checks.v1","dq.health.v1"],"summary":"A data check for unexplained steps in our own archived series (unexplained_step, warn) was added: 42 → 43 checks. It reads the list at /api/steps; a step there is explained by an entry in this register or by a checked note.","reference":"https://agenticfinancegraph.com/api/steps"},{"id":"cp-2026-10-12-poison-rule","date":"2026-10-12","at":"2026-10-12T22:00:00Z","kind":"detector_rule","comparable_from":null,"metric_keys":["poison_events_24h","poison_wallets_targeted_7d","poison_payers_targeted_share_7d","poison_losses_total","poison_loss_usd_total","poison_loss_wallets"],"definition_ids":["poison.events.24h.v2","poison.targeted.7d.v2","poison.targeted.share.7d.v2","poison.losses.v2","poison.loss.usd.v2","poison.loss.wallets.v2"],"summary":"Poisoning rule v2: a zero-value or other-token plant now counts when the look-alike matches a payee on the first 3 and last 4 hex characters (the threshold of Tsuchiya et al., USENIX Security 2025), not only the first 4 and last 4. 1,438 stored zero-value plants on 539 wallets were re-matched (3 of them in the 24 hours before the change; in the 7 days before it they fell on 38 wallets, some already counted as targeted). No new loss followed.","reference":"https://agenticfinancegraph.com/def/poison.events.24h.v2"},{"id":"cp-2026-10-12-domain-scope","date":"2026-10-12","at":"2026-10-12T22:16:00Z","kind":"method","comparable_from":"2026-10-13","metric_keys":["probe_domain_checked","probe_domain_verified_rate","probe_domain_file_rate"],"definition_ids":["probe.domain.checked.v1","probe.domain.verified.v1","probe.domain.file.v1"],"summary":"Endpoint-domain verification stops counting profile links on shared public platforms (github.com, x.com and similar) and reserved example names, where no agent can publish a proof. Only the first sweep (21:57 UTC on 12 October) included them.","reference":"https://agenticfinancegraph.com/def/probe.domain.checked.v1"}]}