Regulatory Report Submission Cycle (Published Edit Checks)

Note (dated 2026-08-04): a chain composes independent computations performed in an analytical order. No step's output is carried into the next step by the execution model; each step computes from its own declared inputs. This is true of every chain on this site, including chains whose per-step copy below reads as a data handoff.

Reading mode (RUNBOOK): Run these steps in order. A hard-rule failure in the edit check actually stops the chain before a submission bundle is produced.

Two-step pre-submission cycle over a report instance and a freely published rule set. Step 1 evaluates the instance against policy-supplied validity and quality edits, keyed by the published edit identifier, and records which deactivated rules were suppressed so a stood-down rule cannot produce a false failure. Step 2 computes period-over-period variance per line item against policy-supplied materiality thresholds and flags which movements require a written explanation. Step 1 is gated hold on any hard-rule failure, and in that state the cycle stops before a submission bundle is produced. Step 2 is gated dual_control(2) for filing release, satisfied by a preparer and a distinct signing-officer approver; a flagged material movement carrying no explanation annotation holds the release for review, and the submitter identity is recorded separately at transmission. The bundle is the filer's own evidence and does not satisfy any issuer's submission requirement.

OpenChainGraph · 2 Steps regulatory reporting Hash-Anchored §4 chain_depth:2 Client-Side · Zero PII

Chain Stages · 2 Steps

1 ROOT node hold

Report Edit-Check Runner art-484-regrpt-editcheck-runner

Per-rule pass/fail keyed by the published edit identifier, with failing cell references and computed-versus-reported values, feeds the variance stage

§27 Human Accountability gate: any hard-rule failure puts the cycle on hold and no submission bundle is produced in that state.

MCP Call · run_regrpt_edit_checks

{
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "name": "run_regrpt_edit_checks",
    "arguments": {}
  },
  "id": 1
}

↓

2 TERMINAL node dual_control(2)

Report Variance Explainer art-485-regrpt-variance-explainer

Ranked material movements and the explained/unexplained split close the submission-evidence bundle under dual control. Final stage.

§27 Human Accountability gate: filing release is dual_control(2): a preparer and a distinct signing-officer approver. A flagged material movement with no explanation annotation holds the release for review, and the submitter identity is recorded separately at transmission.

MCP Call · explain_regrpt_variance

{
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "name": "explain_regrpt_variance",
    "arguments": {}
  },
  "id": 1
}

Question: The quarterly report is due at the regulator. Do any published edit checks fail on the figures we are about to submit? Run the AINumbers MCP tool `run_chain` with {"chain":"regrpt-submission-cycle"} and read `art-484-regrpt-editcheck-runner.summary.overall_pass`. Re-run with `inputs` for `art-484-regrpt-editcheck-runner`, reusing the values that result echoed, with `report_instance` changed so its cell RCON2948 value is 750. 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/regrpt-submission-cycle.html

Regulatory Report Submission Cycle (Published Edit Checks)

Ask your agent