
Agent Skills
✓ Text source-checked Nov 9, 2026A standardized way to give AI agents new capabilities and expertise.
Source-checked against the vendor's site · rebuilt from the vendor's own pagesProfile updated Link checked How we rate
In short
Agent Skills is a free, open standard for extending AI agents with specialized knowledge. It uses portable folders to load specific instructions on demand, keeping context usage low.
Agent Skills is a lightweight, open format for extending AI agent capabilities with specialized knowledge and workflows. A skill is a folder containing a SKILL.md file. The SKILL.md file includes metadata such as name and description, along with instructions that tell an agent how to perform a specific task. Skills can also bundle scripts, reference materials, templates, and other resources.
Agent Skills packages procedural knowledge and specific context into portable, version-controlled folders that agents load on demand. Agents load skills through progressive disclosure, starting by loading only the name and description of each available skill at startup. When a task matches a skill’s description, the agent reads the full SKILL.md instructions into context. The agent then follows the instructions, optionally executing bundled code or loading referenced files as needed. Full instructions load only when a task calls for them, allowing agents to keep many skills on hand with a small context footprint.
The Agent Skills format was originally developed by Anthropic, released as an open standard, and has been adopted by a growing number of agent products. The standard is open to contributions from the broader ecosystem. Agent Skills works in any compatible agent, including Claude Code and OpenAI Codex.
According to Agent Skills: Agent Skills is a lightweight, Open format for extending AI agent capabilities with specialized knowledge and workflows. Its pages list this strength: Lightweight and open format for extending agent capabilities.
— The AI Tool Bible editorial team
Pros
- ✅ Lightweight and open format for extending agent capabilities.
- ✅ Portable, version-controlled folders that agents load on demand.
- ✅ Build a skill once and use it across any skills-compatible agent.
- ✅ Progressive disclosure keeps the context footprint small.
- ✅ Open to contributions from the broader ecosystem.
Facts
- Platforms
- cli, vscode-extension (not source-checked)
- Pricing
- free - Free open standard (not source-checked)
- Open source
- Yes (license: Apache-2.0)
- GitHub stars
- 27,236 (checked Nov 13, 2026) (not source-checked)
- Last GitHub push
- Nov 4, 2026 (not source-checked)
- First commit
- Dec 2025 (not source-checked)
- Model used
- Model-agnostic (not source-checked)
- Hacker News
- 3 stories · top 544 points · last Feb 3, 2026 · checked Nov 7, 2026
Where this tool sits
Use cases
Frequently asked
- How much does Agent Skills cost?
- Agent Skills is a free open standard. It is model-agnostic and designed for broad adoption across different agent products without licensing fees.
- Which AI models and agents support Agent Skills?
- The format is model-agnostic and works in any compatible agent. It has been adopted by a growing number of products, including Claude Code and OpenAI Codex.
- How do agents load and use skills?
- Agents use progressive disclosure. They initially load only skill names and descriptions. When a task matches, the agent reads the full SKILL.md instructions and executes bundled code or loads referenced files as needed.
- What is inside a skill folder?
- A skill is a folder containing a SKILL.md file with metadata and instructions. It can also bundle scripts, reference materials, templates, and other resources to support specific tasks.
- Is Agent Skills an open standard?
- Yes, it was originally developed by Anthropic and released as an open standard. The standard is open to contributions from the broader ecosystem to support cross-product reuse.