01 · Headline finding

Of 103 Texas Business Court opinions in the corpus, 63 (61%) resolved on a threshold ground — jurisdictional, procedural, or early-procedurali — without adjudicating an underlying claim on the merits.

The corpus is, 21 months in, predominantly a jurisdictional-doctrine corpus. The split here separates opinions for which a discrete claim adjudication is coded (merits-reaching) from those resolving only on procedural or jurisdictional grounds without reaching the underlying claim (threshold). Seven of the threshold-side opinions carry merits-shaped early-procedural postures (7 Rule 166(g) early-legal-issue rulingsi; 4 Rule 91a dismissalsi); they are coded threshold because the disposition did not adjudicate a claim on its merits. Bucketing convention: threshold = jurisdiction-only postures (remand, special appearance, plea to jurisdiction, removability, transfer) plus Rule 91a, Rule 166(g), and motions to dismiss; merits = every other posture (n=40). Hover any segment for the per-posture breakdown; click to filter the matrix below. The early corpus is a jurisdictional map, not a mature merits corpus.

Threshold rulings — 63 opinions

    Merits-reaching — 40 opinions

      Hover a segment for the breakdown · click to filter the matrix

      Source note

      The Texas Judicial Branch provides short staff descriptions for public convenience. Those descriptions are not part of the opinions, are not official statements of the court, and are not legal authority. Texas Judicial Branch, Business Court Opinions. The Codex uses staff descriptions only for intake; all final coding traces to PDF paragraph pin cites in the opinions themselves. The threshold/merits split reflects a coded ruling_bucket (threshold / merits / mixed) validated against the PDF, not a mechanical inference from an empty claims field.

      Corpus snapshot & data coverage

      Corpus snapshot: 2024 Tex. Bus. 1 through 2026 Tex. Bus. 40, last checked June 15, 2026.

      Official PDF URL: 100%
      Date ISO: 100%
      Page count: 95%
      Fees / damages / AIC: coverage-limited
      Petition-level pleaded claims: pending docket backfill
      Claim taxonomy: discussed / adjudicated in opinion

      No corpus-level inference is drawn from fields with low coverage. Petition-level pleaded-claim coverage requires docket / pleading backfill and is not complete unless an opinion row is marked “pleading verified.” Where a claim count appears, it distinguishes claims discussed or adjudicated in the opinion from pleaded causes of action in the underlying petition; a single opinion may carry multiple values on the former axis.

      Methodology — full taxonomy and coding rules

      Corpus universe. Every opinion published in the Texas Judicial Branch Business Court Opinions repository from October 1, 2024 through the last-corpus-check date above. Both full published opinions and memorandum opinions are included; the dataset distinguishes the two via the memOp flag.

      Threshold vs. merits bucketing. An opinion is coded threshold when the disposition resolves on jurisdiction-only postures (remand, special appearance, plea to the jurisdiction, removability, transfer) or on Rule 91a / Rule 166(g) / motion to dismiss. Every other posture is coded merits. The split is computed live in the browser from the posture field; see bucketFor() in dashboard.js for the exact keyword list.

      Ruling categories. Five buckets are recognized: granted, denied, mixed, remanded, other. The category is derived from the leading token of the ruling field with two fall-throughs (dismissed, substantive ruling) folded into other. The full ruling text is preserved verbatim and displayed in the matrix.

      Primary-source linkage. Every opinion in the matrix carries a pdf_url that resolves directly to txcourts.gov. No CourtListener, Westlaw, Lexis, or other aggregator URL appears in the dataset (primary-sources rule).

      Cell-level filtering. Every visualization element (iceberg segment, Sankey stream, heatmap cell, judge card, division-month cell, ribbon dot) is a click target. Clicking applies a filter to the matrix below and scrolls into view. Filter chips appear above the matrix and clear individually or via the Reset button.

      Coverage caveats. The Court does not publish an order docket with file-date metadata; commencement-date coding for HB 19 § 8 questions traces to the opinion text plus the underlying state-court petition where verified. Petition-level pleaded-claim coding (which differs from the opinion's claim-adjudication coding) is incomplete and only displayed where flagged pleading verified.

      HEADLINE FINDING · AS OF June 15, 2026

      After 21 months of operation, the Texas Business Court has issued 103 substantive opinions across five divisions — but only 39% reach the merits, and 35% of the corpus is concentrated in just two judges (Whitehill, Bullard). Threshold rulings still outnumber merits-reaching opinions in the corpus to date.

      • 103Opinions in corpus
      • 39%Reach the merits
      • 5 of 11Divisions operational
      • 35%Issued by top 2 judges
      • 21 mo.Since Sept. 2024 opening

      Source: SMU CGI hand-coded opinion corpus, 2024 Tex. Bus. 1 through 2026 Tex. Bus. 40, cross-checked against txcourts.gov/businesscourt/opinions. Threshold/merits split derived from per-opinion posture coding (see Glossary § "Threshold vs merits"). Judge concentration tracked in the Judges card grid below.

      02 · From courtroom to outcome

      Division → procedural posture → ruling

      Every published opinion traces a single path. It enters at one of the five operational divisions (1st · Dallas, 3d · Austin, 4th · San Antonio, 8th · Fort Worth, 11th · Houston), encounters a specific procedural posture (motion to remandi, plea to the jurisdictioni, Rule 91ai, summary judgmenti), and exits with a ruling. Stream width tracks opinion volume; stream color encodes the ruling category.

      What to look for

      The corpus's largest single posture-disposition concentration: 19 of 29 remand-posturedi opinions (66%) resulted in the canonical disposition granted. Grants outnumber denials at remand by a wider margin than at any other posture in the corpus. Click any stream to filter the matrix below.

      Streams thinner than two opinions are dimmed for legibility. Rare postures and rulings are grouped under "Other".

      03 · Production over time

      Opinion volume by division and month

      Each cell is the number of opinions a division published in a given month. Darker = more opinions. The heatmap surfaces the corpus's two-track tempo: month-to-month volume is uneven, and the divisions do not stack on the same monthly cycle. Click any cell to filter the matrix to those opinions.

      What to look for

      Volume builds through 2025; the 1st (Dallas) and 11th (Houston) divisions account for the densest cells, and the most recent month on the right is typically the largest as the court hits its stride. Cells with zero opinions appear muted; click any populated cell to drill in.

      Cells colored on a 0–5+ scale; the Court has produced at most a handful of opinions per division per month. Months on the X-axis run from October 2024 to May 2026.

      04.5 · Reader's reference

      Glossary — every term, in Plain English and Academic detail

      Every division, procedural posture, ruling type, and doctrinal anchor that appears in the Sankey above is defined here. Each card opens to a two-tab popup: a Plain-English tab with examples and analogies, and an Academic-detail tab with the black-letter rule, the controlling Texas Business Court opinion, and Bluebook 21st citations to primary sources. Two-tab pattern: SMU CGI, ExxonMobil Stress Tests, exxon-publication.pages.dev/stress_tests.

      Divisions

      The geographic units of the Texas Business Court. The 88th Texas Legislature created eleven divisions in HB 19; five — the 1st, 3d, 4th, 8th, and 11th — received funding and opened for filings on September 1, 2024. Tex. Gov't Code Ann. §§ 25A.002, 25A.003 (West 2025); HB 19, 88th Leg., R.S. (Tex. 2023). Across these five divisions the corpus contains 103 opinions (48 full published; 55 memorandum), October 2024 through June 2026.

      Procedural postures

      The specific motion or vehicle that put each opinion in front of the Court. These are the doctrinal levers — they determine the burden, the standard of review, and what the Court is actually deciding.

      Rulings

      "Granted" and "denied" carry posture-dependent operative consequences. A granted plea to the jurisdiction dismisses the suit for lack of subject-matter jurisdiction as filed; a granted motion to remand sends the action to the state district court from which it was removed. Posture and ruling must be read together. Of 19 granted remand motions in the corpus, 2 rest on § 25A.004(d) monetary-floor reasoning (Clean-Co under (d)(1); OWL Assetco 1 under (d)(2)); the remainder rest on commencement-date bars, counterparty status, or other statutory predicates.

      04.6 · Doctrines at a glance

      Five doctrines, by share of substantive holdings in the first 21 months

      Five doctrinal questions account for the largest share of substantive holdings in the corpus. Each is anchored in a specific statutory text, a controlling Texas Business Court opinion, and a practitioner-facing takeaway. Click any doctrine for plain-English meaning, the black-letter rule, the controlling TBC opinion, and the Bluebook authority.

      04 · What was being argued, in what shape

      Claims discussed or adjudicated × motion type

      Across the merits-reaching opinions, this matrix shows which claims travel with which motion types. Rows are the claims discussed or adjudicated in each opinion (coded from court text, not from petitions); columns are the motions that produced the opinion. Darker cells = more opinions at that intersection. Heatmap counts opinion-claim pairs (an opinion deciding multiple claims contributes once per claim); the merits-reaching tile counts opinions. Pair total ≠ opinion total when an opinion adjudicates more than one claim.

      What to look for

      On the heatmap's coding — claims on which an opinion's disposition actually operates on the merits — no single claim-by-posture cell dominates the early corpus. The largest cell is breach of fiduciary duty × other-merits rulings at 4 opinions, followed by breach of contract × summary judgment at 3. A broader pleaded-claim coding (claims present in the complaint but not necessarily reached on the merits) produces a larger contract footprint and is reported separately in the Methods section. Click any cell to drill into those opinions.

      1 ·

      05 · Doctrine deep dive

      Clean-Co Systems v. Enterprise Products Operating — the springing-aggregation question under § 25A.004(d)(1)

      Hero Box · Why this case matters

      A Texas Business Court judge sent a $688,142 invoice fight back to district court, ruling that two decades of unrelated work under an umbrella service contract cannot be added together to clear the court's $5 million jurisdictional floor.

      Tier 1 · Plain English

      In 2003 a chemical-cleaning company, Clean-Co, signed an umbrella service agreement with an Enterprise Products affiliate. That agreement promised nothing on its own — work was ordered, one purchase order at a time, over the next 22 years. In 2025 Clean-Co performed work on a single purchase order; Enterprise didn't pay; Clean-Co sued in regular district court to collect about $688,000. Enterprise tried to move the case to the Texas Business Court — a specialized court for high-dollar commercial disputes — by arguing that because it had paid Clean-Co more than $7.8 million across all the purchase orders since 2003, the relationship as a whole cleared the court's $5 million jurisdictional floor.

      Judge Grant Dorfman rejected that "add it all up" theory and sent the case back to district court. The point of the $5 million floor, he reasoned, is that the transactions giving rise to the lawsuit must themselves be big enough — not the parties' entire historical relationship. The 22 years of prior work spanned different facilities, different scopes, and possibly different Enterprise entities; only one purchase order was actually in dispute. None of the rest was in controversy. The decision is the first Texas Business Court opinion to reject this kind of "cumulative-payments-under-an-MSA" jurisdictional theory, and it tightens what defendants must show to keep a case in the specialized court.

      Tier 2 · Lawyer's takeaway

      The doctrinal hinge is § 25A.001(14)'s "series of related transactions" element, not the § 25A.004(d)(1) jurisdictional grant standing alone. The court signals a four-factor relatedness test as a matter of law: temporal span, scope congruence, facility/geographic congruence, and counterparty/affiliate continuity — applied against the transactions-in-controversy limitation the plaintiff advanced (P9, P11). The court reaches this result even on the deferential pleading-only standard of C Ten 31, LLC v. Tarbox, 2025 Tex. Bus. 1, ¶¶ 49–50 (3d Div.), and explicitly declines to make the hearsay ruling load-bearing (P8). The open question is what happens when temporal span is short, scope is congruent, and the same affiliated counterparty signs every order — Clean-Co does not foreclose aggregation in that fact pattern; it tightens it.

      For a defendant invoking § 25A.004(d)(1) under MSA-pleading: (i) plead transaction-by-transaction relatedness with specificity, not a global cumulative-payment number; (ii) attach admissible business-records evidence of the cumulative figure (the Gold affidavit's failure is instructive); (iii) be prepared to address affiliate continuity across the MSA-signatory, the order-issuer, and the work-tasker (P7 n.6); (iv) consider pleading single-integrated-transaction in the alternative under Shell W. E&P (P10); and (v) anchor consideration value to time-of-contracting, not cumulative ex post performance, per Reed v. Rook TX, 2025 Tex. Bus. 34, ¶ 17 (P4 n.3).

      In Clean-Co Systems v. Enterprise Products Operating, 2026 Tex. Bus. 32 (Dorfman, J., 11th Div., May 20, 2026), Enterprise argued that a "qualified transaction"i within the meaning of Tex. Gov't Code Ann. § 25A.001(14) (jurisdictional grant at § 25A.004(d)(1)) is established once cumulative payments under a long-running master service agreement clear the $5 million aggregate-value threshold, even when the single Purchase Order in dispute had an original not-to-exceed cap of $154,360 and a now-claimed invoice value of $688,141.87 after scope-of-work changes — an amount the plaintiff itself characterized as "less than one million dollars" (P9). The slider below illustrates how that aggregation would reshape the jurisdictional analysis.

      The doctrinal point

      Judge Dorfman granted remand on one dispositive ground: the historical purchase orders invoked by Enterprise do not constitute a qualifying "series of related transactions" under Tex. Gov't Code § 25A.004(d)(1) because they "span more than two decades, involve differing scopes of work, and were performed at diverse facilities" The Court separately sustained a hearsay objection to Enterprise's Oracle payment records. Judge Dorfman expressly rested the remand on the related-transactions ground; the opinion does not state, and does not depend on, the evidentiary ruling as an alternative ground. The holding is fact-bound rather than categorical. The Court did not announce a per-se rule against aggregation under § 25A.001(14); it held that the MSA history before it did not constitute "a series of related transactions" on the facts pleaded. Based on SMU CGI's survey of all 103 coded Texas Business Court opinions in the SMU CGI canonical corpus as of the most recent build, Clean-Co appears to be the first to reject the cumulative-payments-under-an-MSA aggregation theory advanced under § 25A.001(14)'s "series of related transactions" definition; the opinion itself does not claim first-impression status. The § 25A.004(d)(1) qualified-transaction floor governs this analysis; § 25A.004(d)(2)'s amount-in-controversy threshold is doctrinally separate and not at issue. Of 19 opinions in which the procedural posture is remand and the canonical disposition is granted, only two turn on a § 25A.004(d) monetary floor: Clean-Co (§ 25A.004(d)(1) qualified-transaction floor) and 2025 Tex. Bus. 30, OWL Assetco 1 v. EOG Resources (Bullard, J.) (§ 25A.004(d)(2) amount-in-controversy floor, addressing aggregation of joined-party claims under § 25A.004(i)). The remaining 11 turn on commencement-date bars (HB 19 § 8i), counterparty status (TBC jurisdiction over non-business-court parties), or other statutory predicates.

      This case · single purchase order

      < $1M

      Amount in controversy on the disputed PO. Standing alone, well below § 25A.004(d)(1)'s qualified-transaction threshold — $5 million post-HB 40 (effective Sept. 1, 2025); $10 million under HB 19 for filings predating Sept. 1, 2025. Clean-Co was decided May 20, 2026 under the post-HB 40 regime. The separate § 25A.004(d)(2) amount-in-controversy basis ($5 million for actions arising out of qualified transactions and the other § 25A.004(d) categories per HB 40 §§ 43, 45 (effective Sept. 1, 2025; reduced from $10 million under HB 19)) is not at issue here.

      Cumulative payments under the 2003 MSA (slider)

      $0.0M

      $0 $2.5M $5M $7.8M

      Not a qualified transaction — single PO under $1M, no springing aggregation.

      Court's reasoning

      "[T]he historical purchase orders on which Enterprise relies span more than two decades, involve differing scopes of work, and were performed at diverse facilities — potentially operated by different affiliated entities. Only one such purchase order — the most recent one in the alleged series — is at issue in this case. … [T]he remainder are not in controversy." Clean-Co Sys. v. Enter. Prods. Operating, 2026 Tex. Bus. 32 at P11.

      — Dorfman, J., 2026 Tex. Bus. 32

      Takeaway. On the facts of Clean-Co, Enterprise's pleaded cumulative-payments figure under a multi-decade MSA — approximately $7.8 million across the historical purchase orders (P3, P6), though the Court sustained Clean-Co's hearsay objection to the Oracle-platform affidavit on which that figure rested, leaving the cumulative-payments number unsupported by admitted evidence (P7 & n.5) — did not satisfy the "series of related transactions" requirement of § 25A.001(14) for the under-$1 million purchase order in dispute. The plaintiff's motion to remand was granted. The Court did not foreclose a § 25A.001(14) aggregation theory on different facts.

      06 · Who is writing the doctrine

      The bench at a glance

      Ten Business Court judges have authored published opinions in the corpus's first 21 months. Each card shows that judge's opinion volume, grant rate, share of threshold-only rulings, and doctrinal signature — the procedural posture each judge has authored most often. Counts describe opinion output, not case difficulty, caseload, reversal risk, or judicial quality.

      What to look for

      Whitehill (19) and Bullard (17) are the corpus's highest-output opinion authors; together they account for 36 of 103 opinions (35%). Bullard authored the leading opinions on HB 40 retroactivityi (2025 Tex. Bus. 47); Bullard also authored the AIC-aggregation remand in 2025 Tex. Bus. 30 (OWL Assetco 1 v. EOG Resources) under § 25A.004(d)(2), a separate qualified-transaction question from HB 40 retroactivity. Adrogué authored the leading opinions on the HB 19 § 8 commencement-date bar (2024 Tex. Bus. 8; 2025 Tex. Bus. 2 and 8); Andrews authored the leading opinions on Rule 166(g)i early-legal-issue rulings and on the Marathon Oil force-majeurei arc; Dorfman authored Clean-Co Systems. Click a card to read those opinions.

      Sort by:

      07 · Repeat appearances in published opinions

      Repeat-appearance ribbon

      Fifteen captions account for 40 of 103 opinions (39%). Each horizontal lane is one caption; each dot is one opinion, positioned on a time axis from October 2024 to today and color-coded by ruling. "Caption" denotes a distinct adverse-party pairing as it appears on the face of the opinion; captions are not deduplicated by underlying business dispute.

      The litigation arc

      Primexx Energy Opportunity Fund v. Primexx Energy Corporation generated five opinions across partnership, drag-along, fiduciary-duty, and personal-jurisdiction questions. Slant Operating v. Octane Energy Operating produced four. Marathon Oil v. Mercuria Energy America traces the Winter Storm Uri force-majeure arc across three opinions (two published, one memorandum). A small set of multi-opinion captions accounts for a substantial share of the corpus's published output to date.

      Time axis: October 2024 (court's first published opinion) → today. Lanes ordered by total opinion count descending.

      FIGURE · OPINION-ISSUANCE VELOCITY

      From a slow start to a 16-opinion month.

      Twenty-one months in, the Texas Business Court’s opinion output is still climbing. The cumulative curve marks three inflection points: the first published opinion (Oct 2024), the first 9-opinion month (Dec 2025), and a record 16-opinion May 2026; June 2026 is in progress.

      TBC opinion-issuance velocity, Oct 2024 — June 2026Cumulative opinion count rises from 2 (Oct 2024) to 103 (June 2026). Monthly counts shown as orange bars; cumulative as red line. Dec 2025 was the first 9-opinion month; May 2026 set the record at 16 opinions.0255075100CUMULATIVE OPINIONSPER MONTHOct '24: 2 opinions this monthNov '24: 4 opinions this monthDec '24: 2 opinions this monthJan '25: 3 opinions this monthFeb '25: 5 opinions this monthMar '25: 4 opinions this monthApr '25: 3 opinions this monthMay '25: 7 opinions this monthJun '25: 3 opinions this monthJul '25: 3 opinions this monthAug '25: 6 opinions this monthSep '25: 3 opinions this monthOct '25: 5 opinions this monthNov '25: 4 opinions this monthDec '25: 9 opinions this monthJan '26: 3 opinions this monthFeb '26: 5 opinions this monthMar '26: 6 opinions this monthApr '26: 5 opinions this monthMay '26: 16 opinions this monthJun '26: 5 opinions this month (partial month, in progress)051015Oct '24 — cumulative 2 opinions (+2 this month)Nov '24 — cumulative 6 opinions (+4 this month)Dec '24 — cumulative 8 opinions (+2 this month)Jan '25 — cumulative 11 opinions (+3 this month)Feb '25 — cumulative 16 opinions (+5 this month)Mar '25 — cumulative 20 opinions (+4 this month)Apr '25 — cumulative 23 opinions (+3 this month)May '25 — cumulative 30 opinions (+7 this month)Jun '25 — cumulative 33 opinions (+3 this month)Jul '25 — cumulative 36 opinions (+3 this month)Aug '25 — cumulative 42 opinions (+6 this month)Sep '25 — cumulative 45 opinions (+3 this month)Oct '25 — cumulative 50 opinions (+5 this month)Nov '25 — cumulative 54 opinions (+4 this month)Dec '25 — cumulative 63 opinions (+9 this month)Jan '26 — cumulative 66 opinions (+3 this month)Feb '26 — cumulative 71 opinions (+5 this month)Mar '26 — cumulative 77 opinions (+6 this month)Apr '26 — cumulative 82 opinions (+5 this month)May '26 — cumulative 98 opinions (+16 this month)Jun '26 — cumulative 103 opinions (+5 this month)Oct '24Dec '24Feb '25Apr '25Jun '25Aug '25Oct '25Dec '25Feb '26Apr '26Jun '26Dec 2025:first 9-opinionmonth (prior max 7)May 2026: 16 opinions— largest single monthCORPUS-TO-DATE103 opinions
      Cumulative opinions (left axis, 0–100)Monthly opinion count (right axis, 0–16)

      How to read. The red line is cumulative; the orange bars (lower band) are per-month issuance. May 2026 closed at 16 opinions—the largest single month to date; June 2026 is in progress. Twelve months elapsed before the corpus crossed 30 opinions; the next 30 came in five months; the most recent 30 in five months. A docket maturing as five divisions hit stride.

      08 · Every opinion

      Filter, sort, drill into the full corpus

      All 103 Texas Business Court opinions (48 published, 55 memorandum), citation by citation. Use the filter chips, the search box, or click any visualization above to drill in. Sort by clicking a column header. Export the filtered set as CSV.

      Showing 103 of 103 opinions
      Citation Date Case Div. Judge Posture Ruling Claims / Issues  

      09 · How to cite this work

      How to cite this work

      This is a living research product: the underlying corpus and analysis are versioned and updated as the court issues new opinions. The dataset and the interpretive analysis are co-citable.

      Bluebook

      Shane Goodwin, The Texas Business Court Codex v1.3.0, SMU Corp. Governance Initiative (May 23, 2026), https://smucgi.pages.dev/research/texas-business-court/dashboard/ (last visited [date of access]).

      Companion paper: Shane Goodwin, The Texas Business Court Codex: A Doctrinal Audit of the First Nineteen Months (SMU Corp. Gov. Init. Working Paper No. 2026c, forthcoming SSRN).

      APA

      Goodwin, S. (2026). The Texas Business Court Codex (Version v1.3.0) [Interactive dataset and doctrinal analysis]. SMU Corporate Governance Initiative. https://smucgi.pages.dev/research/texas-business-court/dashboard/

      BibTeX

      @misc{goodwin2026tbccodex,
        author    = {Shane Goodwin},
        title     = {The Texas Business Court Codex},
        year      = {2026},
        version   = {v1.3.0},
        publisher = {SMU Corporate Governance Initiative},
        howpublished = {\url{https://smucgi.pages.dev/research/texas-business-court/dashboard/}},
        note      = {Underlying corpus: current Texas Business Court opinion corpus (counts and breakdown queryable from the SMU CGI canonical opinions.json), October 2024 -- present. Accessed [date].}
      }

      Dataset (Bluebook)

      Shane Goodwin, The TBC Opinions Master Matrix (v1.3.0), SMU Corp. Governance Initiative (May 23, 2026), https://smucgi.pages.dev/research/texas-business-court/dashboard/.

      Contributor roles (CRediT taxonomy)

      Shane Goodwin — Conceptualization · Methodology · Investigation · Data curation · Formal analysis · Software · Visualization · Writing — original draft · Writing — review & editing · Project administration · Supervision. Editorial review — SMU CGI research staff. Citation verification & visualization implementation — research assistance via the Claude agent.

      HOW WE WORK

      Four standing rules behind every cell on this page.

      The Codex is hand-coded against primary sources. The rules below are not aspirational — they are the gate every datum passes through before it lands on this page.

      RULE 01

      Primary sources only

      Every opinion linked to its txcourts.gov PDF; statutes linked to the Texas codified text; cited cases linked to the issuing court’s docket or PDF. No CourtListener, Westlaw, Lexis, or other aggregator URLs in the dataset (cross-referenced in the coverage panel above).

      RULE 02

      Bluebook 21st throughout

      Citations follow Bluebook 21st short-form discipline; pin-cites where the page is available; id. only at the next adjacent reference.

      RULE 03

      Per-opinion human review

      Posture, ruling line, and doctrinal tags are hand-coded by SMU CGI, not auto-extracted. Each row in the matrix is editor-reviewed.

      RULE 04

      Independence from litigants

      SMU CGI takes no funding from parties to TBC litigation; no ghost-author engagements; no advance review of summaries by counsel.