diff --git a/docs/TICKETS.md b/docs/TICKETS.md index 80b72e5..6710e95 100644 --- a/docs/TICKETS.md +++ b/docs/TICKETS.md @@ -77,7 +77,7 @@ Newest first. Update on every open/close. | #000034 | Hessian alignment under φ_linear | open · awaiting go/no-go (#000018 follow-up) | 2026-05-09 | — | | #000033 | Claim-pack pillar VII (combinatorics) | closed · landed 2026-05-09 (live in shard 000.db; lift verified) | 2026-05-09 | — | | #000032 | combinatorics@v1 π* (integer counting kernel) | closed · landed 2026-05-09 | 2026-05-09 | — | -| #000031 | Surface-ingest cited textbooks for claim-pack warrant promotion | in progress · Phase 1 + Phase 2 landed 2026-05-09; **2026-05-10: 92/92 (100%) claim-pack records resolved** across 18 textbook substrates + 54 citation-aliases + cascade tuning; Phase 3 verifier wiring ahead | 2026-05-09 | — | +| #000031 | Surface-ingest cited textbooks for claim-pack warrant promotion | in progress · Phase 1 + 2 + 2.5 (92/92) + 3 (verifier wiring) all landed 2026-05-09 / 2026-05-10; via_citation_alias attribution + source-side title-from-author backfill follow-ups still open | 2026-05-09 | — | | #000030 | Math π* expansion: SymPy substrate (algebra · calculus · linalg) | closed · all 7 phases + 1b landed 2026-05-09 (`abe5988`) | 2026-05-09 | — | | #000029 | Claim-pack source (axiom/theorem JSON bundles) | closed · landed 2026-05-09 | 2026-05-09 | — | | #000028 | Multi-modality witness for canonical shapes | closed · landed 2026-05-09 + follow-ups (capital ledger · sample rate) | 2026-05-08 | — | diff --git a/docs/tickets/ticket-000031-surface-ingest-cited-textbooks.md b/docs/tickets/ticket-000031-surface-ingest-cited-textbooks.md index 9c3f7fd..4aef420 100644 --- a/docs/tickets/ticket-000031-surface-ingest-cited-textbooks.md +++ b/docs/tickets/ticket-000031-surface-ingest-cited-textbooks.md @@ -1,6 +1,6 @@ # Ticket #000031 — Surface-ingest cited textbooks for claim-pack warrant promotion -**Status:** in progress · Phase 1 + Phase 2 landed 2026-05-09 (chunk-resolver, proof writer, 18 derivations rows for Hilbert pillar IV); **2026-05-10 push: all 92/92 claim-pack records resolved** across 18 textbook substrates (PD/CC-BY-SA/CC-BY/GFDL/MIT mix), 54 fox-decided citation-alias rows + 13 term-alias rows, plus cascade tuning in `_build_record_query_cascade` (parenthetical-phrase + OR-fallback variants); Phase 3 (verifier wiring to upgrade audit_mode from ANCHOR → EVIDENCE-WARRANTED on the four-rung ladder) still ahead +**Status:** in progress · Phase 1 + Phase 2 + Phase 3 all landed 2026-05-09 / 2026-05-10. Phase 1 (PD textbook surface ingest) + Phase 2 (chunk-resolver + proof writer + 18 derivations rows for Hilbert pillar IV) landed 2026-05-09. Phase 2.5 (single-day push: 18/92 → 92/92 across 18 textbook substrates + 54 citation-aliases + cascade tuning) landed 2026-05-10. Phase 3 (verifier wiring: warrant_chain_lookup + WARRANT_MISSING suppression when cited chunk has a warrant-resolver derivation row + warrant_proven_claim_idxs render tail) landed 2026-05-10. Two follow-ups still open: via_citation_alias process_id attribution (~15 LOC) + source-side title-from-author backfill in HTML / textbook_tex ingest (~30 LOC). **Opened:** 2026-05-09 **Scope:** Ingest the classical textbooks cited by every record in the claim-pack source (`#000029`) — Mendelson 1997, Enderton 2001, @@ -497,33 +497,42 @@ audit reviewer can re-judge. The schema doesn't distinguish these tiers today; reviewers walk `citation_aliases` table + read `decision_rationale` fields. -### Phase 3 — verifier wiring (NOT YET LANDED) +### Phase 3 — verifier wiring (LANDED 2026-05-10) -The data substrate is now COMPLETE (all 92 records have a -`derivations.proof_blob` row). The four-rung-ladder upgrade -that lifts answers citing claim-pack-records-with-derivations -from ANCHOR-WARRANTED to **EVIDENCE-WARRANTED** still needs to -happen in the verifier. Sketch: +Implemented in commit `a57b194` (~491 LOC + 9 new tests). -- The Q&A verifier currently produces `audit_mode` + `violations` - per answer. The four-rung ladder is computed in - `arborist.cli._render_audit_label._ladder_rung_for_lattice` - from those two. -- For each cited claim-pack record, look up whether it has a - ``derivations.process_id='warrant-resolver-v1'`` row. If yes, - REMOVE any `WARRANT_MISSING` violation it would have triggered - (or add a positive `WARRANT_PROVEN` annotation). -- Display label upgrades to EVIDENCE-WARRANTED automatically per - the existing ladder logic. +- New module `arborist/qa/warrant_chain.py` — + `warrant_chain_lookup(shards_dir)` returns the frozenset of + `core_root` values having `derivations` rows with `process_id + LIKE 'warrant-resolver-v1%'` (matches both `+alias` and base + variants). One sqlite query per Q&A run, walks main shards + + sibling crawl/ shards. Tolerates missing tables. +- `verify_claim_lattice` + `verify_claim_lattice_json` accept + optional `warrant_chain_roots: frozenset[str]` parameter. When + the lexical `warrant_check` fails for a claim AND any cited + evidence's `source_root` is in the set, `WARRANT_MISSING` is + suppressed and the claim_idx flows to a new + `warrant_proven_claim_idxs` verdict field. +- `runner.ask` computes the chain set once from the conn's main + DB directory before verifier invocation. Fail-closed default: + empty set → behavior identical to pre-Phase-3. +- `query.py` threads `warrant_proven_claim_idxs` from verdict + into the result dict. +- `cli.py:_render_warrant_tail` adds `· warrant proven via chain + ×N` segment when `warrant_proven_claim_idxs` is non-empty. + Distinct from the pre-existing source-level + `_render_warrant_chain_tail` (which counts cited SOURCES with + chains; this counts CLAIMS). +- The four-rung ladder rung promotes naturally: no + WARRANT_MISSING + no soft demotes → EVIDENCE-WARRANTED via the + existing `_ladder_rung_for_lattice` logic. No render-layer + ladder change needed. -This is a verifier change — touches well-tested code. Worth its -own ticket so the regression risk is bounded. -6. **Bench** — verify a known case (e.g., Modus Tollens - claim-pack record → Aristotle Prior Analytics chapter) - actually promotes from ANCHOR to EVIDENCE. +Live smoke: `warrant_chain_lookup(~/.arborist/shards)` returns +exactly 92 core_roots (matches 2026-05-10 100% claim-pack +warrant-resolution). -Estimated size: ~300-500 LOC across the resolver + proof writer -+ verifier wiring + bench fixtures. Single substantial commit. +1652 tests pass (was 1643). ### Phase 3 — proprietary-textbook decision (deferred)