Rule-Driven Pipeline from Scraping to Feeds
Clawfeed models each source as rules plus an execution pipeline: inputs are webpages or endpoints and outputs are standard RSS/Atom items. The point is to confine scraping uncertainty to the rule layer so downstream systems only consume stable feed URLs. The pipeline typically includes extraction, normalization, deduplication, and ordering to prevent noisy re-emission across refresh cycles. From an ops standpoint, rule-driven design supports versioning and rollback, making upstream adaptation safer without reshaping the whole system.
