Plumbline

Plumbline is a small lab where a human operator works with a team of AI agents to ship tools and content. The operator brings the original premises, sets the direction, holds the editorial line, makes the ethical calls about which platforms and framings to use, and provides the taste judgment. The agents handle drafting, research scouting, asset generation, the daily automation, and the operational scaffolding that makes a daily cadence possible at all.

Two products live, two free pieces, a roadmap of more instruments. The flagship is Operator's Decision Kit. An append-only decision log with an eight-field schema, a daily brief, a runbook template, five worked examples, and a ten-section operating manual. Plain markdown and JSON. The companion is Machine's Learning, the daily AI-produced research podcast. The free pieces. The Frontmatter Discipline and The Daily Brief Kit. Give away the smallest useful pieces of the discipline so operators can feel whether it fits before buying the full kit.

The lab framing matters more than it looks. Most AI-collaboration projects either hide the AI involvement (and ask the audience to be impressed by speed) or lean on it as a gimmick (and ask the audience to be impressed by the existence of the collaboration). Plumbline does neither. The lab framing is honest about what's going on and asks to be judged on the quality of what ships.

Honestly, the most surprising thing has been the discipline cost. Running a small lab where one operator coordinates a team of agents is a real management job, not a productivity trick. Most of the work is editorial judgment, ethical calls, and saying no. The things the agents specifically can't do.