Pipeline fill

Source and import new prospects from Sales Navigator or CSV, then confirm they landed — end to end on the vruum CLI and via MCP.

Pipeline fill sources new prospects, researches them, and lands the qualified ones in a segment. This recipe shows the import surfaces — Sales Navigator and CSV — plus a single manual add, then confirms the result in the pipeline. The pipeline-fill skill orchestrates the full source → research → gate → import motion; the csv-pipeline-fill, sales-nav-deep-fill, and yc-pipeline-fill variants pick a specific source.

CLI flow

Import from a Sales Navigator list or saved search, or from a mapped CSV.

# Import from a Sales Navigator list…
vruum people import sales-nav --list <list_id> --limit 25

# …or from a Sales Navigator saved search.
vruum people import sales-nav --saved-search <saved_search_id> --limit 25

# Import from a CSV, mapping the email column to the person's email.
vruum people import csv ./prospects.csv --map email=<column> --campaign <campaign_id>

# Add one prospect by hand.
vruum people create --name "<full name>" --email "<email>"

Pipeline fill itself is a skill that orchestrates these import commands — it is not a single CLI verb, so reach for vruum people import … (above) or the vruum skills entry point (below), not a one-shot command.

MCP flow

Add a prospect directly with create_manual_person, then confirm the segment's shape with inspect_pipeline. Both are default-loaded MCP tools.

create_manual_person(name="<full name>", email="<email>")
inspect_pipeline()                              → segment counts + recent adds

The bulk Sales-Nav and CSV import endpoints are not in the default tool set — drive those from the CLI above, or reach them from MCP through vruum_exec.

Use the skill

The pipeline-fill skill is the source-agnostic orchestrator: it picks a source per segment, runs harness deep research, applies the pre-filter gate, and saves through the backend's authoritative match_score >= 70 gate. Use a variant (csv-pipeline-fill, sales-nav-deep-fill, yc-pipeline-fill) to pin the source.

# CLI: read the packaged playbook.
vruum skills list
vruum skills show pipeline-fill

# MCP: invoke it inline.
invoke_skill(skill_id_or_name="pipeline-fill")