Every field below maps to a real im_scraper.runner flag.
Advanced Options
Products are always collected; --max-products caps them.
Checkpointing is unconditional in the runner, and the backend has no pause
control, so this UI offers Stop instead.
Live progress
No run selected.
Runner log (tail)
—
Export
Excel, JSON and the phone report are produced by the backend writers;
the dashboard only serves those files.
CPU / RAM (scraper's own process-tree measurement)
Data quality (backend run summary values)
Metric definitions from the backend
Business report
(read-only analysis of this run's result.json; every
metric is OBSERVED, DERIVED by the printed rule, or NOT_AVAILABLE)
Business Leads
(entities with contact evidence actually collected;
ordering is by evidence breadth — never a lead-quality score)
Audience
(profiles from the selected run's bounded batch; follower
counts are audience signals, not unique customers)
Product Market Matrix
(products x platforms with only the prices/MOQs/locations
their own records carried)
Competitor Comparison
(select businesses; rows describe platforms, locations and
evidence counts — no ranking, no scores)
Business entities
(cross-platform resolution over all runs; records merge only on shared
phone, website domain, GSTIN or name+city evidence — never on name alone)