- Fix NameError: len(meaningful) → len(meaningful_mature) (ligne 624)
- _auto_synthesize_knowledge() appelé même si pas assez de trades matures,
pour que le Super Contexte se mette à jour à chaque cycle (gate 6h suffit)
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- database.py: add delete_ai_report(), delete_reasoning_state(), delete_kb_entry()
- reasoning.py: DELETE /api/reasoning/reports/{id}
- knowledge.py: DELETE /api/knowledge/history/{id} and /entries/{id}
- useApi.ts: useDeleteAiReport, useDeleteReasoningState, useDeleteKbEntry hooks
- RapportIA.tsx: trash icon on hover in archived reports sidebar
- SuperContexte.tsx: trash icon on hover for history versions and KB entries;
both propagate onDelete through CategorySection down to KbEntry
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- knowledge.py: classify trades by maturity before building synthesis
prompt; only mature trades (≥35% elapsed) contribute to P&L stats
and conclusions; immature trades listed for transparency only
- Add 6h staleness gate on POST /synthesize (force=true to override)
- System prompt now includes hard timing rule: GPT-4o must not revise
existing conclusions because of newly-added immature trades
- useApi.ts: useSynthesizeKnowledge accepts force boolean param
- SuperContexte.tsx: shows amber notice with age when skipped + offers
"Force quand même" button; success banner uses new response shape
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- Add ScorePillars: compact 4-pillar mini-bars always visible under score
bar (📰/📅/📈/⚖️ with color-coded fill — instant visual of pillar distribution)
- Add ScoreJustification: replaces old per-click BucketBreakdown with full
text reasoning shown at once — key_catalyst highlighted in yellow callout,
each pillar shows score + GPT-4o comment + all sub-pillars with comments
- Rename expand button to "↳ Justification du score" for clarity
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
- ai_analyzer: add explicit calibration rules to SYSTEM_SCORER and batch
prompt so GPT-4o produces a spread of scores rather than defaulting
to 50 for all patterns (0-news patterns capped at 35, contra patterns
at 40, high-signal patterns can reach 70-85)
- auto_cycle: add _auto_synthesize_knowledge() called after each auto
portfolio snapshot; skips if last synthesis < 6h old to avoid
redundant GPT-4o calls — Super Contexte now updates automatically
every cycle without manual intervention
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
