A curated collection of Skills, MCP Servers, and Agents for enhancing AI agent capabilities across the Kilo ecosystem—including Kilo Code (VS Code extension), Kilo CLI, and compatible AI agents.
The Kilo Marketplace is a community-driven repository of agent tooling prompts and configurations. It provides three types of resources that extend what AI agents can do:
| Resource | Description |
|---|---|
| Skills | Modular workflows and domain expertise that teach agents how to perform specific tasks |
| MCP Servers | Standardized integrations that connect agents to external tools and services |
| Agents | Specialized agent configurations for focused tasks and workflows |
Skills are self-contained packages that extend an agent's capabilities with specialized knowledge and repeatable workflows. At their core, a skill is a folder containing a SKILL.md file with metadata and instructions that tell an agent how to perform a specific task.
Skills follow the open Agent Skills specification, making them interoperable across any compatible AI agent—not just Kilo.
Key benefits:
- Self-documenting: Easy to read, audit, and improve
- Interoperable: Works across any agent implementing the Agent Skills spec
- Extensible: Can include scripts, templates, and reference materials
- Shareable: Portable between projects and developers
Each skill is a folder containing a SKILL.md file with YAML frontmatter:
skill-name/
├── SKILL.md # Required: Skill instructions and metadata
├── scripts/ # Optional: Helper scripts
├── references/ # Optional: Documentation
├── assets/ # Optional: Templates, resources
└── examples/ # Optional: Example files
Basic Skill Template:
---
name: my-skill-name
description: A clear description of what this skill does and when to use it.
---
# My Skill Name
Detailed description of the skill's purpose and capabilities.
## When to Use This Skill
- Use case 1
- Use case 2
- Use case 3
## Instructions
[Detailed instructions for the agent on how to execute this skill]
## Examples
[Real-world examples showing the skill in action]Best Practices:
- Focus on specific, repeatable tasks
- Include clear examples and edge cases
- Write instructions for the agent, not end users
- Document prerequisites and dependencies
- Include error handling guidance
MCP (Model Context Protocol) is a standardized communication protocol that allows AI agents to interact with external tools and services. Think of it as a universal adapter—any compatible agent can connect to any MCP server to access its functionality.
MCP servers provide capabilities like:
- File system access
- Database queries
- API integrations
- External service connections
How it works:
- The AI agent (client) connects to MCP servers
- Each server provides specific capabilities
- The agent uses these capabilities through a standardized interface
- Communication occurs via JSON-RPC 2.0 messages
MCP servers can run locally on your machine or remotely as cloud services, depending on security requirements.
Browse available MCP servers in the mcps/ directory.
Agents are specialized configurations that define focused roles, workflows, and instructions for tasks such as documentation, code review, and testing.
Browse available agents in the agents/ directory. The modes/ marketplace remains available only for legacy 5.x clients.
We welcome contributions! Please read our Contributing Guidelines for details on:
- How to submit new skills or MCP servers
- Quality standards
- Pull request process
- Code of conduct
- Ensure your contribution is based on a real use case
- Check for duplicates in existing resources
- Follow the appropriate structure template
- Test your contribution across platforms
- Submit a pull request with clear documentation
This repository is licensed under the Apache License 2.0.