Agent skills
Packaged know-how that teaches an agent to do a job well: SKILL.md folders, skill libraries and how teams share them.
What it is
Agent skills are packaged instructions, scripts and reference files that teach an agent to do a specific job well, such as producing a branded slide deck, following a review checklist or running a deployment. Anthropic introduced the Agent Skills format in October 2025: each skill is a folder with a SKILL.md file that the agent loads only when the task calls for it, and other agent tools have since adopted similar formats.
Why it matters to leaders
Skills are how an organisation captures its own way of working, its templates, policies and quality bar, so every agent applies them consistently. They are also code: a skill can run scripts, so it needs the same review and provenance checks as any software you install.
What to watch
Skill libraries from vendors and the community, versioning and sharing across teams, and security review of third-party skills before they reach production agents.
Latest agent skills radars
Questions and answers
What is an agent skill?
A reusable package, usually a folder with a SKILL.md file plus any scripts or templates, that teaches an AI agent how to carry out a particular task the way you want it done.
How are skills different from prompts or MCP servers?
A prompt is a one-off instruction. A skill is a reusable package of instructions and files that the agent loads when needed. An MCP server exposes live tools and data over a protocol; skills often teach the agent how to use those tools well.
Are third-party agent skills safe to use?
Treat them like any third-party code: check the author, read the instructions and scripts, pin a version and test in a sandbox before giving them access to real systems.
Other beats in the village
- Quantum computingQubits, error correction, post-quantum security and the open-source SDKs where quantum developers are spending their time.
- Agentic AIAI that plans, uses tools and acts across several steps: the projects, research and governance questions behind it.
- AI solutionsThe applications built on top of models: retrieval, copilots, document AI and the open-source building blocks behind them.
- AI harnessesThe software around a model that turns it into a working agent: coding agents, terminals, sandboxes and the loop that runs them.
- MCP serversConnectors that let any agent use a tool or data source through the Model Context Protocol, and how to adopt them safely.
- Agentic GitHubThe most-starred agent projects on GitHub and the releases they ship each week: a health check for the frameworks everyone builds on.