# productagent.dev

> productagent.dev is a free, public library of 8 Claude Code harnesses for product managers and operators: skills, agents, shell scripts on a schedule, a permissions fence and a Telegram bot. The flagship, Loops, lets an agent keep working while you sleep: each loop has a goal, a budget, a cadence and a model, runs on your own Mac and Claude or ChatGPT plan inside a fence, and leaves four plain lines per loop in the morning. Free, no paid tier, nothing phones home.

Install Loops (macOS): `curl -fsSL https://productagent.dev/install.sh | bash`
Install one skill: copy its folder into ~/.claude/skills/ (agents go in ~/.claude/agents/), then restart Claude Code.

## Start here

- [About](https://productagent.dev/about): what productagent.dev is, who it is for, how to install, FAQ
- [Loops concept page](https://productagent.dev/harnesses/loops/README.md): the four knobs, five states and the morning report
- [Loops machine](https://productagent.dev/harnesses/loops/MACHINE.md): the runner, the fence and the morning judge
- [Why loops](https://productagent.dev/why): sourced quotes from the people building Claude Code

## Harnesses

- [flow-design](https://productagent.dev/harnesses/flow-design): A progressive design interview: one question per screen, ending in a clickable HTML prototype of a linear user journey (onboarding, checkout, wizard).
- [loops](https://productagent.dev/harnesses/loops): Give an agent a goal, a budget, a cadence and a model. It works one bounded run at a time while you sleep, inside a fence, and hands you four plain sentences in the morning.
- [prototype-swarm](https://productagent.dev/harnesses/prototype-swarm): Crawls a multi-screen HTML prototype for undefined click destinations and spawns one generator agent per missing screen, with a handoff packet written by the source screen, until the click graph is covered.
- [self-maintenance](https://productagent.dev/harnesses/self-maintenance): The one way an unattended run may touch its own loop registry, and the one place a "need from you" becomes a decision you settle with a single command instead of a paste.
- [pm-strategist](https://productagent.dev/harnesses/pm-strategist): A product and strategy advisor agent that blends three lenses (build taste, business structure, PM execution) and pressure-tests ideas instead of cheerleading.
- [todo](https://productagent.dev/harnesses/todo): A TODO list for PMs in one markdown file, reachable from Claude Code and from Telegram: capture, see what's due, triage, and hand items to an agent that asks before it acts.
- [rev-intel-harness](https://github.com/harrisonhjohnson/rev-intel-harness): A CSV of companies in, the people out: point your existing AI subscriptions at target accounts. Open source, MIT.
- [karma](https://github.com/harrisonhjohnson/karma): A personal knowledge graph over everything the loops write: seeds in, relationships out, so a month of night reports stays searchable. Open source, MIT.

## Source

- [GitHub repository](https://github.com/harrisonhjohnson/productagent): the site builds from this repo; every harness has a harness.json manifest
- [Sitemap](https://productagent.dev/sitemap.xml)