Lead research & content pipeline
The job
Research a prospect, score it against a fit rubric, draft outreach or content from a real artifact, and publish only after a human approves.
Why it's hard without a system
The pain you recognize
- Lead research lives in a founder's head or scattered docs — the moment you delegate, the rubric is gone.
- Drafting outreach and content by hand does not scale, but auto-publishing AI output without review destroys trust.
- A spreadsheet is not a system: it cannot run the research, hold the rubric, and gate the publish in one place.
How a team of agents does it
Division of labor, with a human checkpoint
The unit is a team, not a single agent — and the human checkpoint is a product surface, not an afterthought.
Research + qualify
A sales agent researches an account, scores it against a fit rubric stored in an agent-owned table, and records the result. Prospects and content live in agent-owned tables — the system of record, with no external CRM to sync.
Draft
The agent drafts outreach or content from a real artifact — grounded in the research it just did and the table it just wrote to.
Founder approval
The workflow is designed so publishing side-effects pause for a human approval before anything is posted or sent. The founder reviews the draft, then the publish step runs.
What's involved
Plugins and platform
Platform capabilities
- Agent-owned tables — prospects and a content log
- Approval gate before any publish
Approval point
The founder reviews the draft before the publish step runs — nothing is posted or sent without a human sign-off.
Proof surface
The agent-owned tables (prospects, content log), the approval request with the draft attached, and the audit trail of what was published.