{
  "schema_version": "1.0",
  "data_version": "2026.10.22.1",
  "generated_at": "2026-10-22T21:14:39Z",
  "taxonomy_version": "1.1",
  "companies": [
    {
      "company": "Strategy Inc.",
      "ticker": "MSTR",
      "coverage_window": "2024-09-17 → 2026-09-22 (claims dated Sep 2024 → Sep 2026)",
      "edgar_cik": "1050446",
      "claims_count": 26,
      "counts": {
        "result": {
          "hit": 10,
          "partial": 4,
          "open": 6,
          "miss": 2,
          "unscored": 4
        },
        "actionability": {
          "actionable": 21,
          "cautious": 3,
          "promotional": 2
        },
        "materiality": {
          "high": 24,
          "medium": 2
        },
        "venue_class": {
          "sec_filing": 21,
          "earnings_call": 2,
          "executive_social": 2,
          "formal_press_release": 1
        },
        "claim_class": {
          "transactional": 12,
          "strategic": 14
        },
        "source_class": {
          "primary": 24,
          "secondary": 2
        }
      },
      "score": {
        "basis": "resolved ACTIONABLE claims only (taxonomy v1.1 scoring rules); the score object IS the decomposition — no composite scalar is published, by design",
        "execution_trust": {
          "n": 8,
          "hit": 8,
          "partial": 0,
          "miss": 0,
          "hit_rate": 1.0,
          "hit_or_partial_rate": 1.0
        },
        "strategic_trust": {
          "n": 8,
          "hit": 2,
          "partial": 4,
          "miss": 2,
          "hit_rate": 0.25,
          "hit_or_partial_rate": 0.75
        },
        "high_materiality_trust": {
          "n": 16,
          "hit": 10,
          "partial": 4,
          "miss": 2,
          "hit_rate": 0.625,
          "hit_or_partial_rate": 0.875
        },
        "all_scored": {
          "n": 16,
          "hit": 10,
          "partial": 4,
          "miss": 2,
          "hit_rate": 0.625,
          "hit_or_partial_rate": 0.875
        }
      },
      "clarity": {
        "actionable": 21,
        "cautious": 3,
        "promotional": 2,
        "actionable_per_substantive": 0.875,
        "promotional_per_actionable": 0.0952
      },
      "published_headline_read": {
        "trust": "10/16 HIT (63%); 14/16 hit-or-partial (88%)",
        "clarity_counts_a_c_p": "21 / 3 / 2",
        "signal_note": "Financing execution near-perfect; both misses were BTC Yield/$ Gain target raises, revised honestly. Promotional \"never sell\" claims MISSED → $1.25B monetization program (contradiction trigger).",
        "source": "claim-ledger/README.md"
      },
      "published_output_vector_v11": "1. Execution trust — 8/8 = 100% (resolved ACTIONABLE TRANSACTIONAL: rows 1,3,5,8,9,13,15,21 — all HIT). Small-n; every announced financing and every dividend payment closed as promised. 2. Strategic trust — 2/8 = 25% outright; 6/8 = 75% hit-or-partial (resolved ACTIONABLE STRATEGIC: rows 2,4,7,11,16,18,19,20; HIT on 4, 20; PARTIAL on 2, 7, 18, 19; MISS on 11, 16). 3. High-materiality trust — 10/16 = 62.5%; hit-or-partial 14/16 = 87.5% (all 16 resolved ACTIONABLE rows are high-materiality, so this is the union of legs 1+2). 4. Clarity — ACTIONABLE 21/26 (81%), CAUTIOUS 3/26 (12%), PROMOTIONAL 2/26 (8%). Venue mix: formal PR 18, earnings call/presentation 3, SEC filing 3, executive social 2 — commitments live overwhelmingly outside SEC filings (only 3/26 claims originate there). Density: actionable/substantive 21/24 = 0.88; promotional-or-vague per actionable 2/21 ≈ 0.10 — high specificity, low fluff in the claim base. 5. Hype/divergence — high. Two promotional exec-social claims (rows 22, 23; the never-sell doctrine) were both contradicted by subsequent filed/disclosed actions: 32 BTC sold May 26–31 2026, $1.25B Board-authorized BTC monetization program (Jun 29 2026), $218.4M BTC sales YTD per Q2 2026 PR. Promotional intensity peaked (Feb 2026) three months before the pivot to sales — narrative ran inverse to disclosed trajectory. 6. Contradiction flag — YES (mandatory human review). Saylor's \"not going to be selling… every quarter, forever\" (CNBC, Feb 10 2026) and the never-sell X doctrine materially conflict with filed facts: BTC sales executed and a standing BTC-monetization authorization. The admission happened first on the May 5 2026 call (row 24), then the actions followed — so the contradiction is flagged, not gated as hidden: the reversal was disclosed, just directly at odds with the prior identity claim.",
      "source_class_rules": "Per-record source_class is assigned by documented rule from each ledger's method note: MSTR was built from direct EDGAR fetches (primary; executive-social rows secondary). ASST filing-derived rows rest on filing-text reproductions (mirror_verbatim); its press releases were read as company releases (primary); social claims were verified via secondary reporting (secondary). NAKA rests on filing-derived mirrors and relayed releases (mirror_verbatim; rows the ledger itself marks 'secondary relay', and social rows, are secondary). KULR press releases are GlobeNewswire company releases (primary); its SEC-filing rows rest on verbatim filing mirrors (mirror_verbatim) except the two sec.gov-hosted Item 7.01 8-Ks of 2025-07-09 and 2025-10-09 (primary); earnings-call rows rest on transcript aggregators (secondary).",
      "data_file": "/data/claims/mstr.json"
    },
    {
      "company": "Strive, Inc.",
      "ticker": "ASST",
      "coverage_window": "2025-05-07 → 2026-09-22",
      "edgar_cik": null,
      "claims_count": 26,
      "counts": {
        "result": {
          "unresolvable": 5,
          "hit": 14,
          "partial": 4,
          "miss": 1,
          "open": 2
        },
        "actionability": {
          "cautious": 8,
          "actionable": 14,
          "promotional": 4
        },
        "materiality": {
          "high": 15,
          "medium": 9,
          "low": 2
        },
        "venue_class": {
          "formal_press_release": 25,
          "executive_social": 1
        },
        "claim_class": {
          "strategic": 12,
          "transactional": 14
        },
        "source_class": {
          "primary": 25,
          "secondary": 1
        }
      },
      "score": {
        "basis": "resolved ACTIONABLE claims only (taxonomy v1.1 scoring rules); the score object IS the decomposition — no composite scalar is published, by design",
        "execution_trust": {
          "n": 11,
          "hit": 9,
          "partial": 2,
          "miss": 0,
          "hit_rate": 0.8182,
          "hit_or_partial_rate": 1.0
        },
        "strategic_trust": {
          "n": 1,
          "hit": 0,
          "partial": 0,
          "miss": 1,
          "hit_rate": 0.0,
          "hit_or_partial_rate": 0.0
        },
        "high_materiality_trust": {
          "n": 8,
          "hit": 6,
          "partial": 2,
          "miss": 0,
          "hit_rate": 0.75,
          "hit_or_partial_rate": 1.0
        },
        "all_scored": {
          "n": 12,
          "hit": 9,
          "partial": 2,
          "miss": 1,
          "hit_rate": 0.75,
          "hit_or_partial_rate": 0.9167
        }
      },
      "clarity": {
        "actionable": 14,
        "cautious": 8,
        "promotional": 4,
        "actionable_per_substantive": 0.6364,
        "promotional_per_actionable": 0.2857
      },
      "published_headline_read": {
        "trust": "9/12 HIT (75%); high-mat 6/8 HIT, 8/8 hit-or-partial",
        "clarity_counts_a_c_p": "14 / 8 / 4",
        "signal_note": "Trust the machinery, discount the adjectives. Venue skew: 25/26 claims via PR, zero via filings or calls — a filings-only screen sees nothing.",
        "source": "claim-ledger/README.md"
      },
      "published_output_vector_v11": "1. Execution trust (resolved ACTIONABLE TRANSACTIONAL): 9/11 full HIT (81.8%); 11/11 hit-or-partial (100%); 0 outright MISS. n=11. (HIT: rows 3, 9, 12, 13, 18, 20, 21, 22, 24; PARTIAL: row 7 ~54% of $1.5B potential, row 26.) 2. Strategic trust (resolved ACTIONABLE STRATEGIC): 0/1 (0%). n=1. The sole resolved actionable strategic claim is row 15 — Q1 2026 BTC Yield guided >15%, actual 11.1% (MISS). All other strategic claims are OPEN (rows 16, 23) or unresolvable by design (CAUTIOUS: rows 1, 5, 6, 8, 10; PROMOTIONAL: rows 4, 11, 14, 25). 3. High-materiality trust (resolved ACTIONABLE, high materiality, both types): 6/8 full HIT (75%); 8/8 hit-or-partial (100%); 0 MISS. n=8 — all 8 are transactional (rows 3, 7, 9, 12, 13, 20, 21, 26). No resolved actionable strategic high-materiality claims exist: rows 1 and 6 are CAUTIOUS (excluded), row 23 is OPEN. 4. Clarity + density: Venue × actionability — PR: 13 actionable / 8 cautious / 4 promotional (25); executive social: 1/0/0 (1); SEC filings: 0; earnings calls: 0. Actionable/substantive density = 14/22 (64%). Promotional-or-vague per actionable = 12/14 (0.86). Claim volume by venue: PR 96% (25/26), X 4% (1/26), SEC 0%, calls 0%. Resolution-authority mix on the 11 scored transactional rows: 6 at level 1 (SEC filing), 4 at level 3 (formal PR — all flagged where filing evidence should exist), 1 mixed (row 26). Diagnostic, not a discount: Strive commits by press release (25/26 claims) and confirms by filing (6/11 scored resolutions cite SEC evidence) — under v1.0's assumptions a filings-only screen would have seen nothing. 5. Hype/divergence: LOW. One executive-social claim in the window (row 26, Cole X, June 2026, actionable +$2.1B×2 ATM capacity); 0 contradicted by later filings/actions. The social claim was substantive, not promotional, and resolved PARTIAL on verifiability grounds. The hype-venue signal is that a $4.2B capacity expansion originated on X rather than via filing. Cole's Sept 2026 X posts entered the v1.0 research as confirmations, not standalone forward claims — they do not appear in the ledger. 6. Contradiction flag: NO. No material contradiction between promotion and filings/actions. Watching item: row 1's \"minimally dilutive\" framing vs the 1.28B-share resale registration + ongoing ATM — logged UNRESOLVABLE in v1.0 (no quantity/deadline) and stays UNRESOLVABLE; it becomes a contradiction candidate only if the company re-states the framing against current issuance.",
      "source_class_rules": "Per-record source_class is assigned by documented rule from each ledger's method note: MSTR was built from direct EDGAR fetches (primary; executive-social rows secondary). ASST filing-derived rows rest on filing-text reproductions (mirror_verbatim); its press releases were read as company releases (primary); social claims were verified via secondary reporting (secondary). NAKA rests on filing-derived mirrors and relayed releases (mirror_verbatim; rows the ledger itself marks 'secondary relay', and social rows, are secondary). KULR press releases are GlobeNewswire company releases (primary); its SEC-filing rows rest on verbatim filing mirrors (mirror_verbatim) except the two sec.gov-hosted Item 7.01 8-Ks of 2025-07-09 and 2025-10-09 (primary); earnings-call rows rest on transcript aggregators (secondary).",
      "data_file": "/data/claims/asst.json"
    },
    {
      "company": "Nakamoto Inc. (formerly Kindly MD)",
      "ticker": "NAKA",
      "coverage_window": "2025-05-12 → 2026-09-22",
      "edgar_cik": null,
      "claims_count": 25,
      "counts": {
        "result": {
          "hit": 9,
          "partial": 5,
          "open": 2,
          "unresolvable": 1,
          "unscored": 8
        },
        "actionability": {
          "actionable": 15,
          "cautious": 2,
          "promotional": 8
        },
        "materiality": {
          "medium": 8,
          "high": 12,
          "low": 5
        },
        "venue_class": {
          "formal_press_release": 7,
          "sec_filing": 8,
          "executive_social": 10
        },
        "claim_class": {
          "transactional": 14,
          "strategic": 3,
          "null": 8
        },
        "source_class": {
          "mirror_verbatim": 15,
          "secondary": 10
        }
      },
      "score": {
        "basis": "resolved ACTIONABLE claims only (taxonomy v1.1 scoring rules); the score object IS the decomposition — no composite scalar is published, by design",
        "execution_trust": {
          "n": 12,
          "hit": 9,
          "partial": 3,
          "miss": 0,
          "hit_rate": 0.75,
          "hit_or_partial_rate": 1.0
        },
        "strategic_trust": {
          "n": 2,
          "hit": 0,
          "partial": 2,
          "miss": 0,
          "hit_rate": 0.0,
          "hit_or_partial_rate": 1.0
        },
        "high_materiality_trust": {
          "n": 10,
          "hit": 5,
          "partial": 5,
          "miss": 0,
          "hit_rate": 0.5,
          "hit_or_partial_rate": 1.0
        },
        "all_scored": {
          "n": 14,
          "hit": 9,
          "partial": 5,
          "miss": 0,
          "hit_rate": 0.6429,
          "hit_or_partial_rate": 1.0
        }
      },
      "clarity": {
        "actionable": 15,
        "cautious": 2,
        "promotional": 8,
        "actionable_per_substantive": 0.8824,
        "promotional_per_actionable": 0.5333
      },
      "published_headline_read": {
        "trust": "12.0/14 (86%); high-mat-weighted 8.0/10 (80%)",
        "clarity_counts_a_c_p": "17 / 2 / 10",
        "signal_note": "Headline flatters: all 10 HITs transactional; all 4 PARTIALs quantity/strategy. 78% of Bailey's social output promotional. Widest narrative-to-substance divergence.",
        "source": "claim-ledger/README.md"
      },
      "published_output_vector_v11": "Scoring: HIT = 1, PARTIAL = 0.5. Resolved ACTIONABLE rows = 14 (rows 9, 13 OPEN excluded; row 14 UNRESOLVABLE excluded). 1. Execution trust (resolved ACTIONABLE TRANSACTIONAL): rows 1,2,3,4,5,6,7,8,10,11,12,15 → 9 HIT + 3 PARTIAL → 10.5 / 12 = 87.5% (n=12, small) 2. Strategic trust (resolved ACTIONABLE STRATEGIC): rows 16, 17 → 0 HIT + 2 PARTIAL → 1.0 / 2 = 50% (n=2, very small — directional only) 3. High-materiality trust (resolved ACTIONABLE, high materiality, both types): rows 2,3,4,5,8,10,11,12,16,17 → 5 HIT + 5 PARTIAL → 7.5 / 10 = 75% (n=10) 4. Clarity (+ density): ACTIONABLE 15 / CAUTIOUS 2 / PROMOTIONAL 8 (table rows 1–25). - SEC filing: 6A / 2C / 0P. Formal press release: 7A / 0C / 3P. Executive social: 2A / 0C / 5P. - Density: 15 actionable substantive claims; (8P+2C)/15A = 0.67 promotional-or-vague claims per actionable claim; claim volume by venue: filing 8, PR 10, social 7. - Social promotional share: 5/7 table rows = 71% (78% incl. the split-out \"Bitcoin Yield\" sub-claim from the v1.0 pass) — the v1.0 \"78% of Bailey's social output is promotional\" finding survives. 5. Hype/divergence: unchanged from v1.0 — maximalist accumulator narrative on X (\"One Nakamoto = One million Bitcoin,\" \"$1.0mm/btc is the floor,\" \"no bear market for several years\") vs filings showing a net BTC seller in every quarter since the $679M launch buy, a four-lender debt rollover chain, a 1-for-40 reverse split, and a 99% stock decline. Widest narrative-to-substance divergence of the four companies in this exercise. 6. Contradiction flag: YES — human review required. Evidence: (a) rows 18–19 (\"1 Nakamoto = 1m Bitcoin\") vs filed holdings 5,764.91 → 4,467 BTC and falling every quarter (Aug 19 2025 PR; 10-Qs through Q2 2026); (b) row 16's below-NAV sell-BTC-to-buy-back-shares framing vs BTC actually sold to repay lenders with no evidenced 2026-program buyback execution (Q2 2026 10-Q; Jun 11 2026 8-K); (c) row 25 (\"no bear market for several years\") vs BTC ~$126k (Oct 2025) → ~$60k (Aug 2026). Per v1.1 scoring rule 6, material promotion-vs-filings contradiction is a hard gate, not a trust adjustment.",
      "source_class_rules": "Per-record source_class is assigned by documented rule from each ledger's method note: MSTR was built from direct EDGAR fetches (primary; executive-social rows secondary). ASST filing-derived rows rest on filing-text reproductions (mirror_verbatim); its press releases were read as company releases (primary); social claims were verified via secondary reporting (secondary). NAKA rests on filing-derived mirrors and relayed releases (mirror_verbatim; rows the ledger itself marks 'secondary relay', and social rows, are secondary). KULR press releases are GlobeNewswire company releases (primary); its SEC-filing rows rest on verbatim filing mirrors (mirror_verbatim) except the two sec.gov-hosted Item 7.01 8-Ks of 2025-07-09 and 2025-10-09 (primary); earnings-call rows rest on transcript aggregators (secondary).",
      "data_file": "/data/claims/naka.json"
    },
    {
      "company": "KULR Technology Group, Inc.",
      "ticker": "KULR",
      "coverage_window": "2024-12-04 → 2026-09-22",
      "edgar_cik": null,
      "claims_count": 21,
      "counts": {
        "result": {
          "partial": 6,
          "miss": 4,
          "hit": 7,
          "unresolvable": 2,
          "open": 1,
          "disclosure": 1
        },
        "actionability": {
          "cautious": 10,
          "promotional": 4,
          "actionable": 6,
          "null": 1
        },
        "materiality": {
          "high": 14,
          "low": 1,
          "medium": 6
        },
        "venue_class": {
          "formal_press_release": 10,
          "sec_filing": 5,
          "earnings_call": 6
        },
        "claim_class": {
          "strategic": 15,
          "transactional": 5,
          "null": 1
        },
        "source_class": {
          "primary": 12,
          "mirror_verbatim": 3,
          "secondary": 6
        }
      },
      "score": {
        "basis": "resolved ACTIONABLE claims only (taxonomy v1.1 scoring rules); the score object IS the decomposition — no composite scalar is published, by design",
        "execution_trust": {
          "n": 2,
          "hit": 2,
          "partial": 0,
          "miss": 0,
          "hit_rate": 1.0,
          "hit_or_partial_rate": 1.0
        },
        "strategic_trust": {
          "n": 2,
          "hit": 1,
          "partial": 0,
          "miss": 1,
          "hit_rate": 0.5,
          "hit_or_partial_rate": 0.5
        },
        "high_materiality_trust": {
          "n": 3,
          "hit": 3,
          "partial": 0,
          "miss": 0,
          "hit_rate": 1.0,
          "hit_or_partial_rate": 1.0
        },
        "all_scored": {
          "n": 4,
          "hit": 3,
          "partial": 0,
          "miss": 1,
          "hit_rate": 0.75,
          "hit_or_partial_rate": 0.75
        }
      },
      "clarity": {
        "actionable": 6,
        "cautious": 10,
        "promotional": 4,
        "actionable_per_substantive": 0.375,
        "promotional_per_actionable": 0.6667
      },
      "published_headline_read": {
        "trust": "3/4 (75%); high-mat 3/3 (100%)",
        "clarity_counts_a_c_p": "small-n",
        "signal_note": "Every resolved MISS was a promotional payoff claim; every HIT a mechanical commitment. Filings telegraphed the unwind before the final 8-K.",
        "source": "claim-ledger/README.md"
      },
      "published_output_vector_v11": "1. Execution trust (resolved ACTIONABLE TRANSACTIONAL): 2/2 = 100% (#9, #20 HIT). Small-n — two observations only. 2. Strategic trust (resolved ACTIONABLE STRATEGIC): 1/2 = 50% (#19 HIT; #12 MISS). Small-n. 3. High-materiality trust (resolved ACTIONABLE high-materiality, both types): 3/3 = 100% (#9, #19, #20 HIT; #12 is medium). Small-n. 4. Clarity (+ density): - ACTIONABLE by venue: SEC 1 (#14) · PR 2 (#9, #12) · call 3 (#18, #19, #20) · social 0 → 6 - CAUTIOUS: SEC 3 (#6, #11, #15) · PR 5 (#1, #3, #4, #7, #10) · call 2 (#16, #17) → 10 - PROMOTIONAL: SEC 0 · PR 3 (#2, #5, #13) · call 1 (#8) · social 0 → 4 - Density: actionable/substantive = 6/16 (37.5%); promotional-per-actionable = 4/6 ≈ 0.67; claim volume by venue: SEC 4 / PR 10 / call 6 / social 0 (excl. row #21). - Resolution authority: 14/17 resolved rows anchored at SEC-filing or audited level; ⚠ #4 and #8 resolved on PRs alone, #17 on earnings calls alone. 5. Hype/divergence: zero executive-social forward/promotional claims in the window → no measurable social narrative-to-filing divergence. The promotional intensity that exists lives in PRs and calls (#2, #8, #13) and was deployed while disclosed fundamentals were deteriorating (FY2025 net loss $61.9M; H1 2026 BTC mark-to-market losses ~$31.4M) — a narrative divergence, not a social one. 6. Contradiction flag: NO. No social narrative conflicted with filings; the full reversal (#15's \"long-term holdings\" → #20's authorized full liquidation) was disclosed through filings and pre-announced on earnings calls (#19 May 2026, #20 Aug 2026) *before* the September 8-K made it official; losses were disclosed honestly. None of the hard-gate triggers (material promotion-vs-filing contradiction, repeated high-materiality strategic misses, misleading capital-structure claims) are present. Note: #8 (the \"BTC-First turbocharge\" ambition) is the closest thing to a contradiction with later actions, but as a PROMOTIONAL claim it never enters scored buckets.",
      "source_class_rules": "Per-record source_class is assigned by documented rule from each ledger's method note: MSTR was built from direct EDGAR fetches (primary; executive-social rows secondary). ASST filing-derived rows rest on filing-text reproductions (mirror_verbatim); its press releases were read as company releases (primary); social claims were verified via secondary reporting (secondary). NAKA rests on filing-derived mirrors and relayed releases (mirror_verbatim; rows the ledger itself marks 'secondary relay', and social rows, are secondary). KULR press releases are GlobeNewswire company releases (primary); its SEC-filing rows rest on verbatim filing mirrors (mirror_verbatim) except the two sec.gov-hosted Item 7.01 8-Ks of 2025-07-09 and 2025-10-09 (primary); earnings-call rows rest on transcript aggregators (secondary).",
      "data_file": "/data/claims/kulr.json"
    }
  ]
}