Builds
Every company from the import, and whether its demo exists yet. Jobs are queued here and run on your machine, nothing builds while the runner is off.
Waiting on you
Low confidence first, that is where a wrong guess is most likely to reach a real company. Read the reasoning before you look at the page.
Pipeline
Everything, and who owes the next move.
Assumption scoreboard
How often each axis of the rubric guessed right. Below the sample threshold a rate is noise, so it stays unranked.
Objections heard
Most common gaps
Import leads
A registry scrape. Writes given only, enrichment and inference are
later stages. Existing leads are skipped, never overwritten.
Columns per leads/leads.template.csv. A row with no
website_url is the strongest buying signal in the file.
Call log
After a calling session, export the sheet as CSV and drop it
here. Columns: slug,outcome,notes,objections,next, plus optional
axis:verdict:correction pairs in an assumptions column,
which is what keeps the scoreboard fed.
Send demo
This reaches a real company that did not ask to hear from you. Read it back before you commit.