← All topics

Agents

Designing, evaluating, and shipping autonomous AI agents.

9632 posts

Claude Skill 2

Self-evolving browser automation for Codex, Claude Code, and Cursor—learn reusable site knowledge and safely replay verified routines with a...

Python Apache-2.0 Updated 3w ago
Claude Skill 2

Agent Skill that researches, compares, and shortlists hotels with evidence-backed, like-for-like pricing — prepares the booking decision, ne...

Python MIT Updated 1mo ago
Claude Skill 10

A Skill and CLI tool for Codex and Claude Code that converts images, PDFs, image-based PPTX files, and mixed PPTX files into editable PowerP...

Python MIT Updated 3w ago
Claude Skill 1

Point it at any Grafana → full AI/LLM observability: FinOps by sovereignty, agents, EU AI Act — generated, deployed, vision-verified. Zero-d...

Python MIT Updated 1mo ago
Claude Skill 2

GMOメイクショップが開発・公開するエージェントスキル集

HTML MIT Updated 1mo ago
Claude Skill 1

AI agent skill for staged long-task processing with cross-session resume, cross-client project handoff, and silent chat auto-archive to GitH...

Python MIT Updated 1mo ago
Claude Skill 1

AI Agent Skill (Claude Code) to build, tailor & audit a job-winning resume/CV and LinkedIn profile — optimized for ATS parsing (Workday, Gre...

Python MIT Updated 1mo ago
Claude Skill 2

Claude Code skill that stops prompt-cache thrash: -88% token burn in image-heavy agent sessions, with an audit script that shows exactly whi...

JavaScript MIT Updated 1mo ago
Claude Skill 2

AI-powered Claude Code skills that turn design ideas or Figma files into structured specifications, then generate production-ready UI code.

JavaScript Apache-2.0 Updated 1mo ago
Claude Skill 2

Generate, verify, and maintain AI agent guidance for your OSS project — skills, plugins, and AGENTS.md for Claude Code, Cursor, Codex, Copil...

Rust MIT Updated 3w ago
Claude Skill 4

Format text outputs into ADHD-friendly styles to improve focus and readability for any user.

Python MIT Updated 4w ago
Claude Skill

Skills for Claude Code & claude.ai — token-efficiency: measured 30-50% shorter responses at 100% assertion pass rate, with always-on hooks a...

Python MIT Updated 1mo ago