discover-legal/BigLaw
56.6
Adequate · 16 November 2026
31.1k
lines of production code
Go
with Python, TypeScript
1
measurement over time
What this system is
Behavioural changes
1 commit (0 fixes) modifying eval/proofrun, python
A change to existing behaviour in eval/proofrun, python — 1 commit, 7 files: “v2 PR4b: memo workflow end to end, durable runs, REST API, embedded UI”.
eval/proofrun, python · medium confidence · unverified
10 commits (1 fix) modifying eval
A change to existing behaviour in eval — 10 commits (1 fix), 5 files: “eval/bench: review round 2 (judge process group, signal-proof cleanup)”; “eval: shared proof runs skip with the test that drove them; -count=N drives N runs”; “tests: storetest.Open(t, store.OpenPool) replaces the repeated store-opening block (15 sites)”; and 7 more. Also in this area: 4 commits adding/updating tests in eval; 1 commit adding/updating tests in eval.
eval, eval/bench · medium confidence · unverified
11 commits (2 fixes) modifying ui
A change to existing behaviour in ui — 11 commits (2 fixs), 3 files: “Add admin panel with live settings + proper DocuSeal incorporation”; “v2: start BigLaw 2 from an empty tree”; “fix(ui): proxy /reviews (and /bots, /lpm) in the dev server; pin the list to the router”; and 8 more.
ui · medium confidence · unverified
15 commits (7 fixes) modifying internal/store/migrations
A change to existing behaviour in internal/store/migrations — 15 commits (7 fixs), 12 files: “v2 sign-in: review r2 fixes (stateless sign-in, names only for members)”; “v2 PR3c: vision ingest (images and JPEG-scanned PDFs on the vision role)”; “v2 PR3a: documents — ingest, section tree, search, page reader, ladder, file bounds”; and 12 more (refs \#54).
internal/store/migrations · medium confidence · unverified
2 commits (1 fix) modifying eval/tasks
A change to existing behaviour in eval/tasks — 2 commits (1 fix), 35 files: “v2 PR3a: documents — ingest, section tree, search, page reader, ladder, file bounds”; “v2 PR3a r1 fix: bounded reader child, linear .docx text, streamed downloads, bounded section tree”.
eval/tasks · medium confidence · unverified
236 commits (71 fixes) modifying (repo-wide)
A change to existing behaviour in (repo-wide) — 236 commits (71 fixs), 13 files: “v2 lab-improve r2: verifier fixes (G8 list end, serial sweep fallback, per-table header, cover grant)”; “Add payload indexes, recommend()-based recruitment, and RuVector GNN feedback”; “v2: Manage hooks r1 fixes: withheld effects, hook quarantine, exact approvals”; and 230 more (refs \#60, \#55, \#54, \#62).
(repo-wide) · low confidence · unverified
35 commits (8 fixes) modifying ui/src
A change to existing behaviour in ui/src — 35 commits (8 fixs), 5 files: “v2: Manage hooks r1 fixes: withheld effects, hook quarantine, exact approvals”; “Add admin panel with live settings + proper DocuSeal incorporation”; “v2: start BigLaw 2 from an empty tree”; and 32 more.
ui/src · medium confidence · unverified
6 commits (2 fixes) modifying internal/api
A change to existing behaviour in internal/api — 6 commits (2 fixs), 8 files: “Wave 1 usability: upload to a matter, failure reasons, plain-language memo text, smaller UI”; “api tests: one request helper (server.send) and one Manage write route table the session probe also iterates”; “tests: storetest.Open(t, store.OpenPool) replaces the repeated store-opening block (15 sites)”; and 3 more (refs \#62).
internal/auth · medium confidence · unverified
6 commits (3 fixes) modifying internal/store/storetest
A change to existing behaviour in internal/store/storetest — 6 commits (3 fixs), 6 files: “v2 PR1: fix every r1 committee finding (DSN safety, migrations, files, exposure, CI)”; “v2 PR3b r1 fix: no mixed-model vectors, fair fusion for lexical-only docs, E4 measures what runs”; “v2: bench7 fixes: figure reader cost, finish order, judge precision, observability (\#55)”; and 3 more (refs \#55).
internal/store/storetest · medium confidence · unverified
7 commits (3 fixes) modifying ui/src/components
A change to existing behaviour in ui/src/components — 7 commits (3 fixs), 7 files: “v2 PR4b r2: verifier fixes (2 MEDIUM, L-5)”; “v2 PR11: UI (7 screens) with a recorded engine, pulled forward by owner decision 13”; “Review fixes: rolled deadline and reason kept in Limitations, demo provenance on exports”; and 4 more (refs \#62).
ui/src/components · medium confidence · unverified
8 commits (3 fixes) modifying internal
A change to existing behaviour in internal — 8 commits (3 fixs), 47 files: “v2 lab-r3: round-3 benchmark fixes (sweep schema/order/budget, multi-part requirements, crash journal, left-out evidence, raw deadlines)”; “v2 lab-r3 r2: verifier fixes (whole-part requirements, near-join entity/metric, TOC and duplicate skips, skip coverage and disclosure)”; “v2 lab-r3 r6: narrow partOf framings, .docx paragraph starts and list-item windows, check wording and answers, panic usage”; and 5 more. Also in this area: 9 commits (1 fix) modifying internal/agent.
(repo-wide) · medium confidence · unverified
Fixes
11 commits (7 fixes) fixing internal/store
A fix in internal/store — 11 commits (7 fixs), 6 files: “v2 PR1: fix r3 committee findings (URL credential leak, passwd-less uid, leaner checks)”; “v2 PR3a: documents — ingest, section tree, search, page reader, ladder, file bounds”; “v2 PR3a r1 fix: bounded reader child, linear .docx text, streamed downloads, bounded section tree”; and 8 more (refs \#55).
internal/store · medium confidence · unverified
13 commits (7 fixes) fixing internal/model
A fix in internal/model — 13 commits (7 fixs), 11 files: “v2 bench fixes: a subagent's cut-off empty answer fails the tool, not the run”; “v2 PR2: model client, typed errors, Fake, pricing (embedded prices.json + BIGLAW\_PRICES)”; “v2 PR3c r1 fix: transcription disagreement rule, stamped scans, typed vision failures; ceiling 15,600”; and 10 more (refs \#55, \#54).
internal/model · medium confidence · unverified
15 commits (8 fixes) fixing scripts
A fix in scripts — 15 commits (8 fixs), 2 files: “v2 PR3c r4 fix: short text over a page image is mixed (read, text layer kept, only added lines); ceiling 15,650”; “v2 PR4a: owner decision 14, Go prod LOC ceiling 16,100”; “v2 PR11: UI (7 screens) with a recorded engine, pulled forward by owner decision 13”; and 12 more.
scripts · medium confidence · unverified
3 commits (2 fixes) fixing github.com/discover-legal/BigLaw
A fix in github.com/discover-legal/BigLaw — 3 commits (2 fixs), 1 file: “v2 lab-r3: round-3 benchmark fixes (sweep schema/order/budget, multi-part requirements, crash journal, left-out evidence, raw deadlines)”; “v2 lab-r3 r5: typed paging instead of trimming, a bounded text window for requirements, parser and partOf fixes”; “v2 lab-r3 r4: read each requirement in its section of the document's tree and confirm verdicts against the whole section”. Also in this area: 7 commits (1 fix) modifying github.com/discover-legal/BigLaw.
github.com/discover-legal/BigLaw, ui/src/api, ui/src/screens · medium confidence · unverified
31 commits (16 fixes) fixing cmd
A fix in cmd — 31 commits (16 fixs), 5 files: “Add the one-command installer and publish the image to GHCR”; “v2 PR1: fix r3 committee findings (URL credential leak, passwd-less uid, leaner checks)”; “v2 PR3c: vision ingest (images and JPEG-scanned PDFs on the vision role)”; and 28 more (refs \#55, \#54).
cmd · medium confidence · unverified
Test coverage
12 commits adding/updating tests in ui/test; 6 commits adding/updating tests in eval/tests.
Dependencies
15 commits updating dependencies (4 manifests)
A dependency / build maintenance change in (dependencies) — 15 commits (4 fixs), 4 files: “v2: start BigLaw 2 from an empty tree”; “fix(deps): patch postcss, fast-uri, find-my-way security advisories”; “fix(ui): add @types/node so the tsc build step passes”; and 12 more.
(dependencies) · high confidence · unverified
Written by cca.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.
How this codebase got here
Baseline
- First survey — no prior run to compare against. CAI 56.6.
Lenses
- Code Health 71.9
- Architecture 73.5
- Maturity 77.9
- Readiness 40.5
- Security 65.9
- Accessibility 72.9
- Performance 91.4
Changes since last survey
- 300 commits — 174 feature/other, 126 fixes
By area
- (repo) — 71 commits
- internal/engine — 66 commits
- biglaw-go/internal — 54 commits
- (root) — 28 commits
- internal/docs — 18 commits
- ui/src — 12 commits
- internal/api — 10 commits
- internal/agent — 6 commits
- internal/analysis — 5 commits
- internal/model — 5 commits
- biglaw-go/cmd — 4 commits
- eval/bench — 3 commits
- benchmarks/proof-runs — 2 commits
- cmd/biglaw — 2 commits
- eval/proofruns_test.go — 2 commits
- internal/store — 2 commits
- .github/workflows — 1 commit
- biglaw-go/go.sum — 1 commit
- docs/benchmarks.md — 1 commit
- docs/features — 1 commit
Notable commits
- fix: Merge origin/main (PR #35) into fix/ci-and-citation-parse
- fix: Merge origin/main (PR #36) into fix/null-slices-ui
- fix: Merge origin/main (PR #38) into fix/model-failure-errors
- fix: Merge origin/main (PR #39) into fix/model-failure-errors
- fix: Merge origin/main into fix/model-failure-errors
- fix: Merge origin/main into fix/null-slices-ui
- fix: Merge pull request #35 from discover-legal/fix/domain-correctness
- fix: Merge pull request #36 from discover-legal/fix/ci-and-citation-parse
- fix: Merge pull request #37 from discover-legal/fix/null-slices-ui
- fix: Merge pull request #40 from discover-legal/fix/model-failure-errors
- fix: Merge pull request #51 from discover-legal/v2-bench-fixes
- fix: Merge v2-manage (review fixes) into v2-manage-tools
- fix: Merge v2-manage-tools (r2 fixes) into v2-manage-hooks
- fix: Merge v2-manage-tools (r3 fixes) into v2-manage-hooks
- fix: Review fixes: rolled deadline and reason kept in Limitations, demo provenance on exports
- fix: eval/bench: a pinned, reproducible benchmark suite with a fixed results schema
- fix: fix(agents): end evidence quotes at the next labelled line
- fix: fix(agents): recover inline citations and read confidence from its own line
- fix: fix(api): conflict check on POST /clients; audit roster/profile changes
- fix: fix(api,backend): size cap on the sent bytes; routing parity; bounded attach
- …and 280 more
Architecture
- 0 containers · 2 bounded contexts · 0 dependency edges (baseline)
Written by cca.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.
Survey your own repository
discover-legal/BigLaw was measured the same way every project in this corpus was: the same rubric, at a pinned commit, with the result published in full. Point a surveyor at a repository you know and see whether you agree with it.
About this page
- The score is its most recent published measurement, taken on 16 November 2026 at a pinned commit. It is not a live figure and does not change until the project is measured again.
- Measured at commit 068668307feacd1497b9b1bc754d61d8e46c8593 — the exact code this score is about.
- Scored under rubric-2026.11.4 — the same rubric and the same method as every other entry in this index.
- Measured by Canine Code Analyzer (cca.canine.dev) using cca measure-2026.11.84+preprod-d645ca9bbedd.