What happens when you point an AI coding assistant at product management instead of code? Nanopm is an open-source tool that runs a full planning cycle inside Claude Code's terminal with a single command.
The /pm-run command chains five steps: audit, objectives, strategy, roadmap, and PRD (product requirements document). Each step writes a markdown file that the next step reads, so context builds across the pipeline instead of starting fresh each time.
The more interesting piece is persistent memory. Nanopm stores product knowledge in ~/.nanopm/memory/, which means running an audit six months later doesn't start from zero. It picks up where you left off and compares current state against past decisions.
The tool follows the SKILL.md pattern that's been gaining traction in the Claude Code community - structured markdown files that give the AI agent domain-specific knowledge and instructions. It's the same approach used to turn Claude Code into specialized roles like QA engineer or DevOps lead, now applied to product planning.
For solo founders and small teams without a dedicated PM, this fills a real gap. Writing a proper PRD or maintaining a product roadmap is the kind of work that's important but perpetually deprioritized when you're also writing code and talking to customers. Having it automated inside a tool you're already using for development removes enough friction that it might actually get done.
The output is all markdown files, so there's no lock-in. You can review, edit, and version-control everything the same way you would any other documentation.