This paper extends the Munafiq Protocol's surface-depth divergence detection from document files to financial reporting systems and on-chain cryptocurrency disclosures. The same failure mode the Protocol diagnoses in AI systems (a system presenting a compliant surface while the substrate diverges) is the central problem in financial reporting: a filing's reported numbers (zahir) can diverge from the economic reality they claim to represent (batin), and the divergence is structurally detectable from the filing's own internal consistency without requiring external ground truth. The paper proposes Bayyinah al-Maal ("clear evidence of wealth"), a deterministic structural scanner for SEC filings (10-K, 10-Q, 8-K, DEF 14A) and on-chain cryptocurrency data. The scanner applies the architectural principles proven in the Bayyinah document integrity scanner (145 mechanisms, six closed format gauntlets, 1,671 tests as of April 28, 2026): Tier 1 verified findings from deterministic structural checks, Tier 2 structural anomalies warranting review, Tier 3 interpretive signals that never override Tier 1. Detection operates by structural address on XBRL taxonomy elements and blockchain state (Escape A from the No-Escape Theorem), with zero semantic inference in the detection layer. The four-process taxonomy (Aligned, Compliant, Performing, Misaligned) maps onto financial actors with structural precision. Forty Tier 1 mechanism candidates are proposed across four categories for SEC filings (cross-filing consistency, footnote-to-number reconciliation, year-over-year structural drift, and filing metadata anomalies) plus ten on-chain mechanism candidates for cryptocurrency, each scoped to what is byte-deterministic from on-chain state alone. The v1.0 paper proposes the taxonomy; v1.1 will publish worked mechanism contracts with XBRL address pairs, thresholds, and example finding outputs. Three empirical validation plans are pre-registered: retrospective validation against known fraud cases (Enron, Wirecard, Luckin Coffee, FTX), prospective validation against the full EDGAR XBRL corpus with a falsifiable hypothesis (high-finding-rate clusters must correlate with subsequent enforcement actions or restatements), and cryptocurrency validation against the top 100 projects by market capitalization with honest publication of sahih results. Five honest caveats bound the paper's claims: (1) structural consistency does not prove truthfulness (a perfectly fabricated filing scores sahih); (2) XBRL coverage is not universal; (3) the scanner does not predict prices or make investment recommendations; (4) retrospective validation on known fraud is subject to hindsight bias; (5) on-chain mechanisms are scoped to byte-deterministic state (proxy-pattern interpretation and cross-chain attestation are deferred to v1.1). This is the fourth substrate the Munafiq Protocol has been applied to, following AI systems (DOI: 10.5281/zenodo.19700420), document files (DOI: 10.5281/zenodo.19700445), and programming languages (DOI: 10.5281/zenodo.19776584). The five standing principles transfer unchanged. The novel contribution is the formalization of financial integrity as a five-verdict surface (sahih, mushtabih, mukhfi, munafiq, mughlaq), the Tier discipline applied to financial analysis separating structural checks from interpretive signals, and the use of on-chain blockchain data as a cryptographically verified batin for continuous structural monitoring of cryptocurrency project claims.
Arfeen et al. (Wed,) studied this question.