Benjamin_Boenisch/breakpilot-core
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix: skip_web_search Default auf True — 5x schnellere Pipeline

Browse Source 
Anchor-Search (DuckDuckGo + RAG via SDK) verlangsamt Pipeline von
~50 Chunks/min auf ~10 Chunks/min. Anchors (OWASP/NIST-Referenzen)
koennen nachtraeglich in einem Batch-Job befuellt werden.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
Benjamin Admin
2026-04-13 12:26:01 +02:00
parent c02a7bd8a6
commit 756d068b4f
2 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
control-pipeline/api/control_generator_routes.py
View File

@@ -52,7 +52,7 @@ class GenerateRequest(BaseModel):
max_controls: int = 50 max_controls: int = 50
max_chunks: int = 1000 # Default: process max 1000 chunks per job (respects document boundaries) max_chunks: int = 1000 # Default: process max 1000 chunks per job (respects document boundaries)
batch_size: int = 5 batch_size: int = 5
skip_web_search: bool = False skip_web_search: bool = True # Default True — Anchors nachtraeglich batchen
dry_run: bool = False dry_run: bool = False
regulation_filter: Optional[List[str]] = None # Only process these regulation_code prefixes regulation_filter: Optional[List[str]] = None # Only process these regulation_code prefixes
regulation_exclude: Optional[List[str]] = None # Skip these regulation_code prefixes regulation_exclude: Optional[List[str]] = None # Skip these regulation_code prefixes

2
control-pipeline/services/control_generator.py
View File

@@ -489,7 +489,7 @@ class GeneratorConfig(BaseModel):
max_controls: int = 0 # 0 = unlimited (process ALL chunks) max_controls: int = 0 # 0 = unlimited (process ALL chunks)
max_chunks: int = 0 # 0 = unlimited; >0 = stop after N chunks (respects document boundaries) max_chunks: int = 0 # 0 = unlimited; >0 = stop after N chunks (respects document boundaries)
skip_processed: bool = True skip_processed: bool = True
skip_web_search: bool = False skip_web_search: bool = True # Default True — Anchor-Search verlangsamt 5x, nachtraeglich batchen
dry_run: bool = False dry_run: bool = False
existing_job_id: Optional[str] = None # If set, reuse this job instead of creating a new one existing_job_id: Optional[str] = None # If set, reuse this job instead of creating a new one
regulation_filter: Optional[List[str]] = None # Only process chunks matching these regulation_code prefixes regulation_filter: Optional[List[str]] = None # Only process chunks matching these regulation_code prefixes