Summary
Nate B. Jones introduces Open Skills, a public library of reusable agent procedures designed to solve what he calls procedural debt — the growing friction of re-explaining working methods to AI agents across every new session, tool switch, or team member onboarding. The video follows up on his earlier Open Brain project, which addressed memory portability, and targets the layer above memory: how agents know the way you work, not just what you know.
The current release includes 31 skills organized into seven categories, along with seven runbooks — composed sequences of skills that produce reliable, end-to-end outputs. Example runbooks include a creator workflow (voice memo to published web page, chaining transcription, brain dump processing, voice matching, HTML generation, and site publishing) and a release-day briefing pipeline. Individual skills handle granular tasks such as browser QA (actually opening a browser, checking mobile layout, capturing screenshots, and reporting evidence) and personal voice (matching a person’s writing style rather than producing generic AI output).
The core problem Open Skills addresses is fragmentation: teams using both Cursor and Claude Code end up maintaining separate and drifting versions of the same guidance in cursor rules and CLAUDE.md files. Open Skills provides a portable, tool-agnostic format for agent procedures that can be called from any harness. The library is public, and Jones positions it as complementary infrastructure alongside memory systems for anyone building serious multi-agent workflows.
📺 Source: AI News & Strategy Daily | Nate B Jones · Published June 19, 2026
🏷️ Format: Hands On Build
