Three-step BSA/AML lookback closure cycle over the re-screened population, the resulting disposition sampling frame, and the caller's own per-sampled-item disposition records. Step 1 reconciles re-screening completeness against the declared population, excluding any period with no preserved list snapshot as unverifiable rather than screening it against today's list. Step 2 builds a deterministic disposition sampling frame and reviewer workload allocation over the verifiable population. Step 3 rolls up disposition-coverage, disposition-rationale-presence, and a population-to-sample tie-out into a closed gate value. A hold or escalate outcome routes the cycle to the BSA Officer as reviewer; a completed lookback, including an auto_pass rollup, still requires a separate approval record naming the BSA Officer or designated compliance role as approver. Customer id, alert id, and transaction reference are salted throughout; no plaintext identifier ever enters the chain. This bundle is the institution's own evidence of its lookback closure basis, not a supervisory determination.
reconcile_lookback_completeness{
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"name": "reconcile_lookback_completeness",
"arguments": {}
},
"id": 1
}
build_disposition_sampling_frame{
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"name": "build_disposition_sampling_frame",
"arguments": {}
},
"id": 1
}
hold or escalate rollup routes the cycle to the BSA Officer as reviewer before it is treated as settled. A completed lookback, including an auto_pass rollup, still requires a separate approval record naming the BSA Officer or designated compliance role as approver (§27.1); automation never substitutes for that named sign-off.roll_up_aml_lookback_disposition{
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"name": "roll_up_aml_lookback_disposition",
"arguments": {}
},
"id": 1
}
Copy this paragraph into Claude, OpenClaw, or any MCP-aware agent to run this whole workflow on the server and check one field that answers a real question.
Question: We owe the monitor a disposition sample for the consent-order lookback. How large must the sample be if we tighten the tolerable deviation rate?
Run the AINumbers MCP tool `run_chain` with {"chain":"aml-lookback-closure-cycle"} and read `art-471-disposition-sampling-frame.sample_size`.
Re-run with `inputs` for `art-471-disposition-sampling-frame`, reusing the values that result echoed, with `tolerable_deviation_rate` set to 2. Compare the same field.
Verify: call `verify_execution_hash` (https://mcp.ainumbers.co/mcp) with `claimed_hash` set to `composite_execution_hash` and the full `composite_artifact`.
Ledger, for a human re-check: https://ledger.ainumbers.co/
PII rule: send synthetic or anonymised inputs only. The MCP server runs these kernels and logs no payloads.
Chain page: https://ainumbers.co/chaingraph/chains/aml-lookback-closure-cycle.html