humanlayer/skills

This page introduces an external open-source repository. It is not an HDATF product.

What it does

A set of Claude Code skills from HumanLayer. They rewrite CLAUDE.md files, narrow React prop types, build iterated coding agent workflows on GitHub Actions, design agentic control loops, and explain topics with diagrams and HTML pages.

How it helps ATF

Its control loop and iterated agent loop skills are worth comparing with how Harness runs a task, keeps memory and checks results. The CLAUDE.md rewrite skill could be a reference for instruction files given to executors.

License

MIT Permissive. Commercial use and changes are allowed if the copyright notice is kept.

More in this category

• larashero3-dotcom/lieflat-charts
  Lieflat Charts is an Agent Skill in SKILL.md format that turns data into interactive HTML charts in several editorial styles. When a report is explicitly requested, it builds a full-page HTML report from 12 Chinese and English templates.
• tt-a1i/archify
  Archify is an agent skill for Cursor, Claude Code, Codex CLI and OpenCode. Agents write typed JSON, which Archify checks and compiles into self-contained HTML and SVG diagrams, including before and after views of architecture changes.
• yanliudesign/mono-color-skill
  An agent skill that turns a theme, sentence, object or photo into an original editorial print image for posters, zines, portraits and packaging. It defaults to a controlled two-ink look and keeps pure one-ink when asked.
• s1dashu/ip-as-logo-skill
  An Agent Skill that guides image generation of very simple, rounded mascot logos. It limits shapes and colors, proposes three directions, then generates six candidates after the user approves.
• Tencent/teamai-cli
  TeamAI is a CLI that manages a team's skills, rules, MCP settings and knowledge across AI agents such as Claude Code, Codex, Cursor and GitHub Copilot CLI. Admins publish to a shared git repo, and members' sessions pull the updates.

Only repositories in the ranked Trendshift lists are included, and the lists are used only to find candidates. We do not copy their ranks. Descriptions, licenses and star counts come from each GitHub repository. The notes are our own reading. We have not tested these projects, and a place on a trending list does not prove quality.