Compliance Advisor, Drafting Agent und Validator haben nicht geantwortet
weil qwen3.5 standardmaessig im Thinking-Mode laeuft (interne Chain-of-
Thought > 2min Timeout). Keiner der Agenten benoetigt Thinking-Mode —
alle Aufgaben sind Chat/Textgenerierung/JSON-Validierung ohne tiefes
Reasoning. think:false sorgt fuer direkte schnelle Antworten.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Migrate queryRAG from klausur-service GET to bp-core-rag-service POST with
multi-collection support. Each of the 18 ScopeDocumentType now gets a
type-specific RAG collection and optimized search query instead of the
generic fallback. Vendor-compliance contract review now uses LLM + RAG
for real analysis with mock fallback on error.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Add legal context enrichment from Qdrant vector corpus to the two
highest-priority modules (Requirements AI assistant and DSFA drafting
engine).
Go SDK:
- Add SearchCollection() with collection override + whitelist validation
- Refactor Search() to delegate to shared searchInternal()
Python backend:
- New ComplianceRAGClient proxying POST /sdk/v1/rag/search (error-tolerant)
- AI assistant: enrich interpret_requirement() and suggest_controls() with RAG
- Requirements API: add ?include_legal_context=true query parameter
Admin (Next.js):
- Extract shared queryRAG() utility from chat route
- Inject RAG legal context into v1 and v2 draft pipelines
Tests for all three layers (Go, Python, TypeScript shared utility).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Drafting Engine: 7-module pipeline with narrative tags, allowed facts governance,
PII sanitizer, prose validator with repair loop, hash-based cache, and terminology
guide. v1 fallback via ?v=1 query param.
IACE: Initial AI-Act Conformity Engine with risk classifier, completeness checker,
hazard library, and PostgreSQL store for AI system assessments.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
