The dbn-legal-agent-v3 fine-tune (Track 1 / family) emits a labelled-prose
template — duplicate `answer:` prefixes, markdown-escaped underscores (`\_`),
and a trailing raw JSON blob — rather than the strict JSON the Azure/gpt-4o
path produces via response_format. decodeJsonObject() returned null on that
invalid JSON, so ask() dumped the entire raw blob into `answer`.
Fix at the parse layer (no upstream response_format change, to avoid fighting
the fine-tune's training):
- dbnToolsRepairJsonText(): strip fences, drop only invalid `\_`/`\*` escapes,
then balanced-brace scan collecting every top-level {...} (longest first) to
recover an appended JSON object. Shared by both gateways' decodeJsonObject(),
so all JSON tools benefit.
- dbnToolsParseLabeledFields(): parse labelled-prose into real fields when no
JSON decodes, tolerating escaped key names and collapsing duplicate prefixes.
- ask() null-fallback now builds clean structured fields from the parsed prose
instead of dumping raw; what_remains_uncertain becomes a proper list.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
De-family-ify shared JSON tools (persona-aware routing + neutral base
prompt), make the verification review pick its engine per track
(family/child-welfare -> dbn-legal-agent-v3, others -> gpt-4o interim),
and route product-name strings through dbnToolsProductName(). Rebrand the
MCP/tools surface (mcp.php + i18n mcp_* strings) to Do Better Legal.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Wire the legal-domain persona picker into corpus, deep-research, korrespond and
the dashboard chat. Each endpoint reads the chosen profile, resolves its packages
against client 57, and scopes retrieval via package_ids (falling back to family
when omitted). New dashboard tenants now subscribe to all DBN domain packages so
persona switching survives the subscription intersection.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Generalize the family-locked legal tools into caveauAI persona profiles
(client 57 chat profiles, resolved in-process via the chat_profiles bridge).
Each tool accepts an optional `profile` slug that scopes the corpus package(s),
search method, system prompt and synthesis model; omitting it falls back to the
family-legal package so existing behaviour is unchanged.
- dbnToolsResolvePersona / dbnToolsListPersonas / dbnToolsBootChatProfiles in
bootstrap.php; new api/personas.php + dbn.list_personas MCP tool.
- LegalTools search/ask/corpusContextForSummarize and the BvjAnalyzer /
LegalAnalysis / translate paths take the persona's packages + prompt + model.
- Persona <select> on ask/search/summarize (populated from api/personas.php).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Restores the 3 tools (manifest + invoke arms + invokeExtract helper),
the citation-atom RAG lever in LegalTools/corpus-search, and the catalog
icons. These were live on prod via rsync but uncommitted, so a git-pull
deploy reverted the manifest from 22 to 19 tools.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
- renderTimeline(): group consecutive same-date+actor events into one card
with a bullet list; single events keep their current layout
- Date format: YYYY-MM-DD → "1 Jun 2023" (3-letter month, international)
- Time shown in header when available
- Remove date_type badge; confidence badge replaced by amber ⚠ flag on
low-confidence events only (high/medium border colour still shows)
- LegalTools.php: resolve azure_full/azure_mini to Bedrock Sonnet/Haiku
when DbnBedrockGateway is active; claude_sonnet/claude_haiku also handled
- timeline.php + api/timeline.php: engine labels updated (Claude Haiku/Sonnet);
claude_haiku + claude_sonnet added to valid engine list
- i18n engine labels updated in all 4 languages
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Haiku synthesis had a hard cap of 2500 tokens which truncated the
advocate JSON response (~4-6K tokens), causing decodeJsonObject() to
fail and rendering the raw JSON string as the brief.
Fix: remove the 2500-token Haiku override; introduce a per-mode limit
(6000 for advocate, 4000 for deep-research) in $opts before the engine
branch so all paths benefit.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- expandQueries(): truncate seedDescription to 2000 chars (full uploads were
48K+ tokens, exceeding the 35s timeout with Sonnet); switch to Haiku gateway
when Bedrock is active (fast + adequate for sub-Q generation); timeout → 60s
- interpretSeed(): same Haiku + 60s fix for English non-advocate path
- synthesise(): add explicit azure_mini + Bedrock → Haiku branch so the fast
engine actually uses Haiku (~20-40s) instead of falling through to Sonnet (~180s)
- advocate.php: relabel azure_mini as "Claude Haiku 4.5 (fast)" with accurate
timing; relabel claude_sonnet as "(thorough)" to reflect the distinction
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- DeepResearchAgent: engine guard now accepts dbn_legal_v3, claude_sonnet, claude_haiku
(previously stripped these to azure_mini, breaking dbn_legal_v3 selection and
preventing Claude engines from reaching the correct synthesis branch)
- DbnBedrockGateway: remove response_format=json_object from chat payload — LiteLLM
converts this to a tool-use constraint for Bedrock, routing output into tool_calls
instead of content (root cause of the {} empty brief)
- advocate.php: add Claude Sonnet 4.6 (AWS Bedrock) engine option
- account, billing, dashboard, nav, min-sak: pending UI/flow changes from prior sessions
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Three fixes:
1. bootstrap.php dbnToolsRunLegalCheck(): prepend first 350 chars of synthesis text
to the v3 user message so it validates actual content, not just general law.
2. BvjAnalyzerAgent: fix engine guard — was skipping check for claude_sonnet/haiku;
now skips only when dbn_legal_v3 is the synthesis model (it already IS the check).
3. LegalAnalysisAgent: add post-synthesis dbnToolsRunLegalCheck() call after Pass 3;
add 'legal_check' key to runFullAnalysis() return.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Routes AI tools across three tiers based on task complexity:
- Azure GPT-4o-mini always: redact, translate, timeline-basic, search-legal (mechanical tasks)
- Claude Haiku 4.5 (Bedrock): ask, summarize, timeline-deep, citations (Norwegian nuance)
- Claude Sonnet 4.6 (Bedrock): korrespond, legal-analysis, deep-research, barnevernet-analyze,
discrepancy-find, advocate (public-facing legal output)
No AWS credentials in app — credentials live in LiteLLM on Colin (same as nova-lite).
Rollback: DBN_BEDROCK_ENABLED=false in .env, no code push needed.
Includes extended thinking support for Pro deep-research via chatWithThinking().
Claude Opus 4.7 constant added for future premium tier (needs litellm_config.yaml entry).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
nova-lite ignores json_object constraint and returns {} empty; without
it, it wraps output in ```json fences. Strip fences before decodeJsonObject.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Remove GPU/cuttlefish engine from timeline.php, api/timeline.php, LegalTools.php, tools.js (all 4 langs)
- Add engine-aware credit cost: gpt-4o-mini=1 credit, gpt-4o=2 credits (matches redact pattern)
- Remove multiple attribute from file input (single document only)
- New api/timeline-stream.php: SSE endpoint emitting status events + final result
- New api/timeline-download.php: DOCX export of timeline events
- LegalTools::timeline() gains ?callable $onProgress for live status updates
- tools.js: spinner on run, SSE streaming fetch, Export to Word button
- Save to My Docs was already wired (showSaveResultButton at line 1136)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Remove GPU/regex engine options; keep only azure_mini (1 credit) and azure_full (2 credits)
- Variable credit cost: engine-aware pre-check and charge in api/redact.php; PricingCatalog base = 1
- Fix ATTORNEY not preserved when keepOfficials=true: add to LLM prompt, generic-tag, pseudonym regexes
- Replace Azure credits hint with per-engine credit cost text (all 4 languages)
- Single-file upload only (was: up to 5); simplify status messages
- Clear previous redaction output and show pulsing spinner when a new run starts
- Add "Save to My Docs" button in redact output panel (corpus-save.js path)
- corpus-save.js: capture source_doc_ids from button dataset, pass in POST payload
- api/save-to-corpus.php: accept source_doc_ids, store first as source_url=corpus-doc:{id}
- doc-picker.js: show "✂ Redacted" badge for documents saved from the redact tool
- CSS: .redact-working spinner, doc-item__badge--redact pill styles
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- layout.php: replace large manifesto section (eyebrow/title/sub/4 stats) with
a slim dbn-context-bar strip linking to why-ours, pricing, MCP, and privacy
- i18n.php: add context_bar_* + footer_privacy_link + privacy_page_title keys
to all 4 language blocks (en, no, uk, pl)
- dbn-tools-redesign.css: add .dbn-context-bar styles with mobile wrap
- privacy.php: new standalone page in 4 languages covering in-memory processing,
metadata-only logging, My Case opt-in storage, EU data locations, and HTTPS transport
- footer.php: add Privacy link to first footer column
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Add doc_picker_btn/audio_picker_btn to i18n.php (en/no/uk/pl), add
docPickerBtn to REDACT_I18N + TIMELINE_I18N and audioPickerBtn to
TRANSCRIBE_I18N in tools.js. PHP-render picker labels in legal-analysis,
translate, summarize, tool_form, and layout_footer using dbnToolsT().
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Preserve all existing GET params when building lang switcher URLs so that
switching language on /mcp-tool.php?tool=dbn.redact stays on that tool.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Replace all hardcoded English strings in mcp.php with dbnToolsT() calls
- Add 44 MCP UI chrome translation keys to includes/i18n.php (en/no/uk/pl)
- Generate includes/mcp-tool-translations.php with tool names, descriptions,
and parameter docs translated into Norwegian, Ukrainian, and Polish via Azure OpenAI
- Create mcp-tool.php: parameterized detail page (?tool=dbn.slug) with parameter
table, example request/response JSON, and privacy section, all localized
- Add "View details →" links on tool cards in mcp.php (shown on expand)
- Add translations/mcp-chrome.php and scripts/generate-mcp-translations.php
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
New dbn-tools-redesign.css with warm paper surface, navy tools nav, gold
accent, and per-tool themes via body[data-active-tool]. Updated all 21 tool
PHP pages, shared layout/nav/footer includes, and advocate route (new).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The tool now respects the chosen UI language end-to-end — even if the
source document is Norwegian, a user on EN/UK/PL gets the analysis in
their language. Norwegian statute references (barnevernsloven § 4-25,
EMK Art. 8) and case names (Strand Lobben mot Norge 37283/13) are kept
verbatim because they are proper nouns.
LLM (LegalAnalysisAgent.php):
- extractIssues: prompt asks for question + brief_context in user's
language; statute refs preserved
- answerIssue: Norwegian core system prompt (keeps fine-tune precision)
+ language-coercion line for non-NO; localised context/source labels
- synthesise: overall_assessment, next_steps, disclaimer in user's
language; explicit per-language disclaimer text
- runFullAnalysis empty-case fallback also localised
- what_to_check translated per language
UI:
- 40 new la_* translation keys in i18n.php × 4 languages (NO/EN/UK/PL)
- legal-analysis.php: 4-way lang switcher, dbnToolsT() for every label,
emits window.DBN_LA_I18N for runtime JS strings
- legal-analysis.js: t() helper reads from window.DBN_LA_I18N
- layout_footer.php: emits window.DBN_CURRENT_LANG +
window.DBN_ADDON_I18N so the legal-analysis add-on button works in
the page's language no matter which tool it's invoked from
- tools.js add-on: reads from DBN_ADDON_I18N, passes DBN_CURRENT_LANG
to /api/legal-analysis.php so server responds in same language
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Restores the dbn-legal-agent-v3 fine-tune on ocelot (was silently aliased
to plain qwen2.5:14b in LiteLLM since the viper retirement) and ships a
new tool that uses it via a two-pass flow:
Pass 1 (Azure 4o-mini) → extract up to 5 distinct legal issues
Pass 2 (ocelot v3 only) → answer each issue, ≤350 tokens, with corpus
Pass 3 (Azure 4o-mini) → synthesise overall assessment + next steps
The 12GB-VRAM constraint motivates the split: dbn-legal-agent-v3 stays
hot in VRAM through the 5 sequential per-issue calls because issue
extraction and synthesis run on Azure, not on ocelot.
New surface:
- includes/LegalAnalysisAgent.php
- api/legal-analysis.php (NDJSON streaming endpoint)
- legal-analysis.php (dedicated tool page)
- assets/js/legal-analysis.js (streamed UI with per-issue cards)
- Save-result + case-result.php rendering for legal-analysis output
- Nav registration in all four UI languages
Add-on integration: a "⚖️🇳🇴 Run deep legal analysis on this text"
button now appears on Summarize, Ask, and Redact result pages and
streams the same pipeline inline below the existing result.
Existing tools relabelled: the misleading "🇳🇴 Norwegian specialist v3 ⭐"
option on advocate/deep-research/discrepancy/barnevernet is now honestly
"DBN Legal Agent" — now that the real fine-tune is actually deployed,
the label finally matches reality. The advocate.php v2 option was
removed since the v2 GGUF is retired.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
All tool results can now be saved to My Case manually. Users click
'Save result', type a description, and confirm. This replaces the
previous silent auto-save on barnevernet/timeline/etc., giving users
control over what stays and what it's called (supports multiple runs
of the same tool with different titles).
- CaseResults: extend ELIGIBLE_TOOLS to include summarize, ask, redact,
transcribe; add toolLabel/toolIcon entries; support explicit title
via meta['title'] in save()
- api/case/save-result.php: new client-initiated save endpoint;
accepts tool + title + input_payload + output_payload + meta
- Remove CaseResults::save() auto-save from barnevernet, deep-research,
discrepancy, korrespond, timeline API endpoints
- tools.js: add showSaveResultButton() (exposed as window.dbnShowSaveResultButton);
wire for ask, redact, timeline, transcribe (both file-upload and
stored-audio paths)
- barnevernet.js: wire save button after final result render
- summarize.js: wire save button after renderFinal(); passes sumResults
container so widget appears in the correct #sumResults div
- case-result.php: rich tool-specific rendering for summarize, ask,
redact, transcribe, timeline; update re-run link map to include all
new tools
- tools.css: styles for .save-result-widget and its states (idle,
prompt, done, error)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Forms lacked novalidate and textareas had required, so the browser fired
HTML5 validation before tools.js could intercept — blocking submissions
where text came from the doc picker or file upload rather than the textarea.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Documents saved via save-from-tool or case-upload store content directly
in client_documents.content without being chunked into client_chunks.
dbnToolsFetchDocChunks now falls back to client_documents.content for
any requested doc_ids that returned no rows from client_chunks.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- summarize.php: full custom inline form (replaces tool_form.php wrapper) with
lang switcher, azure_mini/azure_full/gpu engine selector, 8 corpus-slice
toggles (all off by default), doc picker, file upload zone, and textarea
- api/summarize.php: rewritten to streaming NDJSON (matches barnevernet pattern);
accepts JSON payload with text, language, engine, slices[], doc_ids[]
- includes/LegalTools.php: adds corpusContextForSummarize() (keyword search via
ClientRagPipeline) and summarizeWithContext() (engine-aware LLM call with
optional corpus prepend); returns structured JSON matching existing summarize format
- assets/js/summarize.js: self-contained IIFE handling file upload via
api/extract.php, slice toggles, NDJSON stream reader, result rendering,
and trace panel update
- includes/i18n.php: adds 'summarize' to nav in all 4 languages (EN/NO/UK/PL),
inserted after 'redact' in the tool order with icon 'SZ'
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
isPaidUser() was checking DBN_FREE_TIER_BALANCE === undefined, which
is only true for CaveauAI sessions. SSO users (even plus/pro) always
have DBN_FREE_TIER_BALANCE set, so the picker was showing the upgrade
modal for everyone in the SSO flow. Now reads DBN_USER_TIER explicitly.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add "Select from My Docs" button to all text tool forms; free-tier
users see an upgrade modal, paid (CaveauAI) users get a searchable
multi-select modal backed by /api/dashboard/documents.php
- Add "Select from My Audio" picker on Transcribe with single-select
and a "Save to My Audio" button for persisting uploaded clips
- New PHP helpers in bootstrap.php: dbnToolsFetchDocChunks,
dbnToolsClientIdFromSession, dbnToolsInjectDocContent
- timeline, ask, redact APIs prepend selected document content
(fetched from client_chunks SQL) before the textarea text
- api/dashboard/audio-upload.php stores audio files on server and
creates a client_documents row with source_type='audio'
- api/transcribe.php falls back to stored audio via audio_doc_id POST
field when no file is uploaded
- api/dashboard/documents.php supports ?source_type= filter
- tools.js: doc_ids added to JSON payload; stored-audio transcribe path
- New assets/css/doc-picker.css, assets/js/doc-picker.js
- SQL migration: scripts/sql/audio_docs_column.sql
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Relative slug.php URLs resolved to /dashboard/slug.php when the nav
was included from /dashboard/* pages. Prefix with / to make absolute.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add require_once bootstrap.php to all 6 dashboard page files so
dbnToolsT() is available before layout_dashboard.php is included
- Add dash_upload_category_lbl key to no/uk/pl sections of i18n.php
(was only in English); Kategori/Категорія/Kategoria
- Fix broken ternary on upload.php Category label — replace with
dbnToolsT('dash_upload_category_lbl', $uiLang)
- layout_dashboard.php outputs window.DBN_I18N with all js_* keys
so dashboard JS reads locale-aware strings from PHP translations
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- New includes/nav.php: sticky site-wide nav with Tools dropdown, Dashboard
link, compact language switcher, user identity → /account.php, Log out
- New account.php: credits & plan, profile, team, usage sections
- New api/corpus-summary.php: JSON endpoint for corpus doc count + last updated
- Replaces topbar in layout.php, layout_dashboard.php, and dashboard.php
- Fixes hardcoded Norwegian strings in dashboard.php credit cards via dbnToolsT()
- Adds 35 new i18n keys across all 4 languages (en/no/uk/pl) in i18n.php
- CSS: .dbn-nav navbar + .account-* account page styles in tools.css
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
client_users.email has a UNIQUE constraint. SSO users who already have a
CaveauAI account share the same email address, causing provision to fail
with "already belongs to another workspace".
Fix: SSO-provisioned client_users rows use dbn-sso-{client_id}@dbn.tools.internal
as the owner email so they never collide with real account rows. The real
email stays on clients.contact_email for display purposes.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Full private corpus dashboard for tools.dobetternorge.no users — each SSO
account gets an auto-provisioned CaveauAI tenant (clients row, corpus) on
first visit. Includes upload (file/paste/URL), RAG chat with SSE streaming
and citation chips, document CRUD, FalkorDB graph relations tab, and
improved save-from-tool flow with tag/preview support.
- dashboard/{index,documents,document,upload,chat,settings}.php
- api/dashboard/{corpus-init,documents,upload,ingest-status,chat-stream,
save-from-tool,graph}.php
- includes/{CorpusProvision,layout_dashboard,layout_dashboard_footer}.php
- assets/css/dashboard.css assets/js/corpus-save.js (routing upgrade)
- includes/{bootstrap,layout}.php extended for dashboard provisioning
Migration 141 (clients.dbn_sso_uid + import_method enum) applied on chloe.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
With timeout=45s + fallback timeout=30s, the total silence was 75s,
exceeding the 60s H2 idle stream limit in the browser. Remove the
fallback: if dbn-legal-agent-v3 times out or fails, return empty
immediately. Legal check is non-critical (wrapped in try/catch in
generate()); the draft is still correct without it.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
daveadmin