marek/creator
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

update

Browse Source
This commit is contained in:
team3
2026-06-23 16:23:12 +02:00
parent cd1c926599
commit cec8d1ee5f
8 changed files with 114 additions and 119 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
backend/config.py
View File

@@ -42,6 +42,16 @@ RECHERCHE_GRACE = 900
# Prüf-Loops lassen Rest-Beanstandungen danach stehen.
KONSENS_MAX_RUNDEN = 3
# Crawler-Sichtung (Content/Noise) — deterministischer Regel-Filter statt LLM.
# Match: Substring (klein) gegen URL UND Dateiname. Reihenfolge: keep > noise > min_chars > behalten.
# Sonderregeln einfach hier ergänzen.
CRAWL_KEEP_PATTERNS = ["learn-unit", "learn-course"] # immer Content
CRAWL_NOISE_PATTERNS = [ # eindeutig themenfremd → raus
"clubs", "events", "podcasts", "resources", "-u-",
"academy", "pricing", "/plans", "career", "newsletter", "impressum", "login",
]
CRAWL_MIN_CHARS = 400 # zu wenig Text → raus
# Timeouts pro Agenten-Schritt: (Basis-Sekunden, Sekunden pro Baustein/Section).
# Gilt für alle Provider gleich — wer zu langsam ist, wird neu gestartet bzw. überholt.
TIMEOUTS = {