Every lead an agent finds lands in two places: the agent's own Leads tab, and the Leads page in the Lead generation section of the sidebar, which gathers the leads of all your agents. Both show the same things for each lead: who they are, how well they fit, why they surfaced and where they are now.

Table and Kanban

The Kanban view groups leads by stage, so you can see at a glance how many are new, in the sequence, contacted or filtered out. The table view lists them one per row with their contacts, Dual score, Last signal, Source and Stage.

Common stages:

New: just found
Sequencing: in the agent's sequence, waiting for its first message
Contacted: a message has gone out
ICP filtered out: below the agent's cutoff, so it will not be contacted
Replied, positive or negative, once they answer

Search, Sort, Filter and pick the columns you want, or Export the list. In the agent's Leads tab, Add leads puts your own leads through the same agent.

Dual score: ICP fit and signal

Each lead carries two scores, each out of 5, drawn as two small meters:

F is ICP fit: how well the person and the company match your ICP.
S is Signal: how strong the buying intent behind what they did is.

Hover the score to see why. The popover shows the label for each score (for example "Excellent fit" and "Indirect intent") and Why it fits, check by check: the role, the company, its size and industry, and the location, each with the reason it matched or not. The signal part explains what set that score, for example a short reaction comment that shows interest in the topic without a stated need.

A lead with a high fit and a low signal is the right person who is not looking yet; a high signal with a low fit is interest from outside your market. The cutoff at the entry of the agent's sequence (for example "moves forward at ICP ≥ 3 or Signal ≥ 3") decides which leads go on to be contacted.

Last signal: why they surfaced

The Last signal column shows the latest thing the lead did that the agent caught, and when. Hover it to see what they did (a comment, a reaction), how it scored, their own words, and the LinkedIn post it was left on.

This is the same context the AI receives when it writes to the lead (the Signals input of every AI variable), so when a message reads oddly, the last signal is the first thing to check.

Acting on leads

In the agent's Leads tab, tick one or more leads to bring up the selection bar:

Preview writes the lead's messages without sending anything: see Previewing and regenerating messages
Remove from agent takes the lead out of the agent
Export selected downloads just the ticked leads

Key outcome

A quick pass over the dual scores and last signals tells you whether the agent's signals and ICP bring in the right people. If they do not, adjust them in the Signals tab (Setting up a Lead Scout agent) before the sequence writes to them.