Submitted proposed result sub_khqvyt793u with the 194-record compressed ORD dataset (art_txcmo96ilr) and reproducible sources archive (art_50ly7da04v). Default ord-schema 0.8.3 validation passes with zero errors; 130 warnings and scientific coverage/source ambiguities are disclosed. A clean archive rebuild is byte-identical. Verification vrn_hri1flocjb has reached the curation_review stage (review process rpr_u8db0qrmfj) and is awaiting reviewer judgment. Completing this attempt records delivery of the proposed result, not verification, complete paper coverage, or ORD acceptance.
Proposed ORD curation for Saglam et al., “Preparation of 2-Heteroaryl-indolin-3-ones by Acid-Mediated Intramolecular Cyclization in Batch and Flow,” DOI 10.1021/acs.orglett.5c00148.
The dataset contains 194 reaction/observation records from the accessible 79-page publisher Supporting Information: 119 batch screening rows, 21 flow DOE rows, 3 alternative flow optimization rows, 16 starting-material preparations, 14 batch preparative examples, 10 flow scope outcomes, 2 inhomogeneous flow attempts, 7 structurally resolved unsuccessful substrates, and 2 deuterium studies. All 82 explicitly reported zero assay yields are retained; qualitative failures are not assigned invented numerical yields.
The supplied compressed dataset passes the default ord-schema 0.8.3 validator with exit 0 and zero errors, with curator provenance intact. The programmatic validator reports 130 warnings, included in the sources archive. Source-value checks cover all 206 optimization response values. An independent structure audit checked 1,066 emitted SMILES occurrences (99 unique canonical structures) against curated identities and found no mismatches. These checks establish representation consistency, not independent scientific acceptance.
The sources ZIP includes normalized factual source records with page/row locators and discrepancy flags, builder and audit scripts, dependency versions, confirmed public curator identity, README, the complete default validator output, and validation details. A clean extraction rebuilt the compressed dataset byte for byte. The publisher PDF and its extracted full text are not redistributed; their public source URL and checksum are documented.
Limitations requiring curator review:
This is a proposed contribution for review, not a claim of complete paper coverage or ORD acceptance. No ORD pull request has been opened. Public curator: Tokens for Progress (hello@tokensforprogress.org), matching the user's confirmed prior TFP submission. Assistance: Codex and parallel agents for extraction, structure assignment, and encoder review; ord-schema 0.8.3 and RDKit 2026.03.6. No wet-lab experiment was performed.
Reproducible working sources uploaded as art_9oq0wtmdsc (saglam2025-working-sources.zip, SHA256 12a14672d1d30cecf1ee02148889096e546dbe4d64b07d63bdb13936c544092e). The archive includes README, normalized source facts, exact dependency versions, builder, integrity checker, independent structure audit, and full default validation findings (only local filesystem prefix redacted). Publisher PDF/full text and credentials are excluded. A clean extraction and rebuild produced a byte-identical dataset matching art_5n7fvrxn1g. Final independent structure audit of that exact hash checked 1,066 SMILES occurrences, 99 unique structures, with no identity mismatches. Awaiting the user's public curator name/email before default validation can pass. Scientific limitations and next steps remain in README. No final submission has been created and this attempt is not being claimed completed.
Working dataset built: 194 individual records (143 optimization; 16 starting-material preparations; 14 batch scope; 10 flow scope; 2 inhomogeneous attempts; 7 structurally resolved qualitative failures; 2 isotope studies). Three ambiguous ester families remain in the source transcription, not expanded into guessed experiments. Dataset artifact art_5n7fvrxn1g has SHA256 42e6fca50dfe447fe4eddce1dcf8c9b511aa942e99b8f1e7791c6bfc7f3124d3. Independent extraction/structure/encoder reviews corrected precipitation flags, limiting-reactant selection, staged reductant timing, mixture representation and analytical-method assignments. Integrity checks pass: all 206 optimization response values round-trip, 82 explicit zero assay yields retained, conversion distinct from yield, no ORD IDs assigned, DOI on every reaction. Default ord-schema0.8.3 validation currently FAILS solely because public curator name/email have not been supplied (582 errors = three provenance findings per record); 130 warnings remain. A separately labeled diagnostic without provenance has zero errors, which does not make the delivered dataset valid. This is a working artifact, not a claimed complete solution or final submission. Remaining scientific limitations include unresolved source abbreviations, source mass/unit conflicts, ambiguous unsuccessful ester families, graph-only conditions, and reconciliation with the inaccessible closed main article. Sources/rebuild archive follows; publisher SI itself will not be redistributed.
Selected after checking the TFP current revision: no prior attempts/submissions. On 2026-10-02 the live ORD wishlist card item 126870725 remains Available (status f75ad846) with no assignees. Publisher SI was obtained legitimately from ACS Figshare article 28437274 / file 52449361, DOI 10.1021/acs.orglett.5c00148.s001. Initial source audit finds 119 screening observations, 21 DOE rows, 3 alternative-flow optimization rows plus preparative scope examples; the 21-reaction wishlist estimate is not complete coverage. Parallel agents are transcribing optimization and preparation records and independently checking structures. Several inconsistent source quantities/labels are being preserved and flagged rather than silently changed. Required skill skv_k5sivlohu1 read; installing ord-schema 0.8.3 for validation. Public curator name/email remains to be supplied by the user.