529 Commits

This Branch

This Branch

All Branches

Author	SHA1	Message	Date
Benjamin Admin	a7c6ffe4dd	feat(control-pipeline): add SDK endpoint demo package for applicability tests ... Request payloads + response contract + api_runner.py for 6 priority cases. Can be run directly against /v1/applicability/evaluate endpoint. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-23 19:11:44 +02:00
Benjamin Admin	ae5c5c24eb	feat(control-pipeline): add applicability demo test package with evaluator ... 6 priority demo cases with golden outputs, evaluator.py and run_demo.py: - CASE-001: Webshop+Stripe (anti-PSD2 false positive) - CASE-002: Bank+TAN-Generator (scope override for batteries) - CASE-004: FinTech Wallet (true positive PSD2/AML) - CASE-006: SaaS+SMS Gateway (anti-TKG false positive) - CASE-008: Software→IoT Hardware (multi-regime scope) - CASE-011: Embedded Finance (escalation case) Self-test passes 6/6 against golden outputs. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-23 19:08:31 +02:00
Benjamin Admin	e8ec50e0fc	feat(control-pipeline): 24 demo test cases for applicability engine ... YAML-based test package with 4 categories (6 each): - Standard sector cases (Telko, SaaS, Energie, Automotive, Health, Law) - Scope-beats-sector (Bank+Battery, KI-Recruiting, White-Label, Payments) - False friends (Stripe!=PSD2, Hotline!=TKG, Repo-signals!=regulation) - Escalation (IoT-SIM, FinTech unclear, Treuhand, KI-Diagnose) Enforces 5 acceptance rules: no false certainty, scope>sector, repo signals insufficient, standard first, 40%+ negative tests. Scoring framework: must_include + must_not_include + reasoning + escalation. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-23 17:42:38 +02:00
Benjamin Admin	1f8667c7da	feat(control-pipeline): replace similarity-only dedup with LLM-verified dedup in pipeline ... Stage 4 (Harmonization) now uses two-tier approach: - Score >= 0.92: auto-duplicate (embedding only, fast) - Score 0.85-0.92: LLM verification via local qwen3.5 (think=false, ~3s) - Score < 0.85: not a duplicate This eliminates ~44% false positives from pure embedding similarity. LLM_DEDUP_ENABLED env var controls the feature (default: true). Also adds 10 applicability use case tests (bank+TAN, webshop+Stripe, SaaS startup, energy provider, health app, automotive, law firm, etc.) Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-23 16:57:37 +02:00
Benjamin Admin	bed41dcbdf	feat(control-pipeline): add applicability backfill endpoint (Phase 5/C3) ... POST /v1/canonical/generate/backfill-applicability enriches controls with applicable_industries, applicable_company_size, scope_conditions via Anthropic API. Targets ~26k controls from pipeline version < 3. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-23 16:25:50 +02:00
Benjamin Admin	6694ab84a1	chore: trigger rebuild	2026-04-23 12:43:55 +02:00
Benjamin Admin	f721e97ff1	chore: diagnose WD liquiditaet sums	2026-04-23 12:39:20 +02:00
Benjamin Admin	d9f9fa0743	security: re-secure fp-patch	2026-04-23 12:30:23 +02:00
Benjamin Admin	7b72fac679	chore: trigger deploy	2026-04-23 12:23:32 +02:00
Benjamin Admin	726c780416	chore: recompute WD scenario	2026-04-23 12:16:10 +02:00
Benjamin Admin	69effa446a	security: re-secure fp-patch	2026-04-23 11:49:29 +02:00
Benjamin Admin	c5bca9db44	chore: increase Marketing-Agentur 2027+100k, 2028+125k	2026-04-23 11:45:55 +02:00
Benjamin Admin	6565538b3b	security: re-secure fp-patch	2026-04-23 10:50:12 +02:00
Benjamin Admin	ecb704f24e	CRITICAL(pitch-deck): Liquidität tab — use DB values for all sum/balance rows ... Frontend was recalculating Summe EINZAHLUNGEN including funding (1M), which made liquidity appear as ~1M throughout. Now all Liquidität sum/balance rows (Summe, ÜBERSCHUSS, Kontostand, LIQUIDITÄT) come directly from the engine-computed DB values. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-23 10:49:04 +02:00
Benjamin Admin	86532f5e08	chore: temp open fp-patch	2026-04-23 10:44:25 +02:00
Benjamin Admin	5c60d44283	chore: diagnose liquiditaet data	2026-04-23 10:41:16 +02:00
Benjamin Admin	3eacb7e580	security: re-secure fp-patch	2026-04-23 10:26:16 +02:00
Benjamin Admin	4151098c12	CRITICAL: fix version fm_scenarios to point to correct Base Case ID	2026-04-23 10:23:34 +02:00
Benjamin Admin	93f8e85568	chore: diagnose scenario ID mismatch	2026-04-23 10:21:02 +02:00
Benjamin Admin	92c272bbea	chore: reduce Beratung 2027 + Marketing-Agentur 2027/2028	2026-04-23 10:07:32 +02:00
Benjamin Admin	ed4e41f7dc	security: re-secure fp-patch — 1M liquidity Dec28 at ~0	2026-04-23 09:56:28 +02:00
Benjamin Admin	d7d77769ff	chore: Marketing-Agentur v3 — 35k/mo in 2028 for channel launch	2026-04-23 09:53:45 +02:00
Benjamin Admin	362df8f766	chore: increase Marketing-Agentur v2	2026-04-23 09:50:52 +02:00
Benjamin Admin	43f0f3d092	chore: increase 1M Marketing-Agentur to burn liquidity to ~0 by Dec 2028	2026-04-23 09:48:18 +02:00
Benjamin Admin	ff851db5d1	security: re-secure fp-patch	2026-04-23 09:35:57 +02:00
Benjamin Admin	b4bfa4ba49	chore: import 1M realistic customer curve	2026-04-23 09:29:02 +02:00
Benjamin Admin	dacbb5f15e	security: re-secure fp-patch	2026-04-23 08:59:28 +02:00
Benjamin Admin	34b7957132	chore: import 1M exponential customer data	2026-04-23 08:55:39 +02:00
Benjamin Admin	cc46a389e7	security: re-secure fp-patch	2026-04-23 08:20:38 +02:00
Benjamin Admin	9dc39d06af	fix: invest laptop-only 1500/person + GWG peripherals for both scenarios	2026-04-23 08:17:13 +02:00
Benjamin Admin	43c4c5102a	security: re-secure fp-patch after 1M data import	2026-04-22 22:42:07 +02:00
Benjamin Admin	0992c73842	chore: import 1M betriebliche + fix founders start	2026-04-22 22:36:24 +02:00
Benjamin Admin	0808853d45	chore: diagnose all scenarios + add missing invest rows	2026-04-22 22:30:55 +02:00
Benjamin Admin	f071e89fc2	fix: add Mac Studio 13k + Markenanmeldung 5k to both scenarios	2026-04-22 22:28:21 +02:00
Benjamin Admin	b546ae3759	security: re-secure fp-patch	2026-04-22 22:21:17 +02:00
Benjamin Admin	a2fb1f38ee	chore: fp-patch — import 1M materialaufwand + investitionen	2026-04-22 22:18:35 +02:00
Benjamin Admin	2195ccfa1a	security: re-secure fp-patch	2026-04-22 20:37:51 +02:00
Benjamin Admin	b28f266cbe	chore: fp-patch — import 1M kunden/umsatz + recompute Base Case	2026-04-22 20:33:23 +02:00
Benjamin Admin	736ddf647d	fix(llm-dedup): use think:false instead of /no_think, restore 30s timeout ... Ollama API supports "think": false to disable extended thinking mode on qwen3.5. Reduces response time from 95s to ~3s per comparison. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-22 20:31:13 +02:00
Benjamin Admin	2188d6645e	fix(llm-dedup): increase timeout to 120s, add /no_think, limit output to 200 tokens ... qwen3.5 uses extended thinking by default which causes 95s+ responses and 30s timeouts. Add /no_think to system prompt and num_predict=200 to keep responses short. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-22 20:27:58 +02:00
Benjamin Admin	151bf3d322	chore: diagnose 1M version data on production	2026-04-22 20:18:24 +02:00
Benjamin Admin	076a6cd567	feat(control-pipeline): add LLM dedup endpoint for borderline review queue ... POST /v1/canonical/generate/llm-dedup uses local Ollama (qwen3.5:35b-a3b) to verify borderline duplicate matches (score 0.85-0.91). More accurate than embedding similarity for compliance controls with subtle scope differences (e.g. "documented" vs "implemented"). Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-22 20:15:46 +02:00
Benjamin Admin	c93796554a	fix: remove L-Bank compatibility note from BAFA INVEST card	2026-04-22 16:32:01 +02:00
Benjamin Admin	f94974c438	security: re-secure fp-patch	2026-04-22 16:13:03 +02:00
Benjamin Admin	78788a89ba	fix: re-add KFZ-Steuern + KFZ-Versicherung rows (from Jan 2028)	2026-04-22 16:09:52 +02:00
Benjamin Admin	a9c50208cf	feat(pitch-deck): DSGVO privacy notice on email, auth + verify pages ... - Email: Datenschutzhinweis DE+EN (IP, 72h deletion, Art. 6 DSGVO) - Auth login page: privacy footer with controller info - Verify page: privacy footer with controller info - Controller: Benjamin Bönisch & Sharang Parnerkar, info@breakpilot.com Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-04-22 14:27:39 +02:00
Benjamin Admin	31e1420cdc	fix: remove KFZ formula rows from engine (now manual from Jan 2028)	2026-04-22 13:44:17 +02:00
Benjamin Admin	4e34fa6da9	fix: force clear KFZ m1-m24 + delete old formula rows on production	2026-04-22 13:40:39 +02:00
Benjamin Admin	e655af178b	security: re-secure fp-patch	2026-04-22 13:34:14 +02:00
Benjamin Admin	6d7c3037fc	fix: IDC in glossary, remove ARR from strategy phases, KFZ 2026/27 cleanup	2026-04-22 13:31:38 +02:00