Agent Skills Standard#
An open standard (agentskills.io) for packaging reusable capabilities that AI agents can load dynamically. Initiated by anthropic, designed for cross-tool compatibility.
Specification#
A skill is a directory with a SKILL.md file:
skill-name/
├── SKILL.md # Required: YAML frontmatter + markdown instructions
├── scripts/ # Optional: executable code
├── references/ # Optional: documentation
├── assets/ # Optional: templates, resourcesRequired Frontmatter#
name: Max 64 chars, lowercase + hyphens, must match directory namedescription: Max 1024 chars, what it does and when to use it
Optional Frontmatter#
license,compatibility,metadata,allowed-tools(experimental)
Progressive Disclosure#
The key design principle — skills scale without blowing up context:
| Layer | When loaded | Budget |
|---|---|---|
| Metadata (name + description) | Session startup, all skills | ~100 tokens |
| Instructions (SKILL.md body) | When skill activated | <5000 tokens recommended |
| Resources (scripts, references, assets) | When specifically needed | As needed |
Keep SKILL.md under 500 lines. Move detailed reference to separate files.
Implementations#
- claude-code: Primary implementation. Extends the standard with invocation control, subagent execution, and dynamic context injection. Plugin marketplace for distribution.
- kiro: kiro-powers contain skills-like packages (MCP servers + steering files + hooks). Not explicitly Agent Skills standard, but similar pattern.
Comparison with Other Extension Mechanisms#
| Mechanism | Standard | Scope | Loading |
|---|---|---|---|
| Agent Skills | Open (agentskills.io) | Instructions + scripts + resources | Progressive disclosure |
| mcp-protocol | Open | Tool integration with external services | Tool definitions deferred |
| fabric Patterns | Open (MIT) | Curated prompts by task (system.md) | Full pattern per invocation |
| kiro-powers | Proprietary | MCP servers + steering + hooks | On demand |
| Scion plugin-system | Proprietary | Harness + message broker plugins | gRPC |
Agent Skills and MCP are complementary: skills teach the agent how to do things; MCP connects the agent to external tools and data. Fabric Patterns are a simpler predecessor — same concept (curated prompts by task) but without scripts, progressive disclosure, or tool integration. The ten-pillars-agentic-skill-design framework provides comprehensive design methodology for building effective skills.