Skills

更新时间:
复制 MD 格式

Skills package domain expertise into reusable capabilities, each defined by a SKILL.md file containing its description, instructions, and optional supporting files.

Core features

  • Intelligent invocation: The model decides when to invoke a Skill based on your request and the Skill's description.

  • Modular design: Each Skill solves one specific type of task.

  • Flexible extension: Supports custom Skills at both the user and project levels.

Create Skills

Create Skills in one of the following ways:

If names conflict, project-level Skills override user-level Skills.

Method 1: Manual creation

Copy the Skill file to one of the following paths. After restarting your IDE, enter / in the dialog box to view loaded Skills.

Location

Path

Scope

User-level

~/.lingma/skills/{skill-name}/SKILL.md

All projects for the current user

Project-level

.lingma/skills/{skill-name}/SKILL.md

Current project only

Method 2: Automatic creation using built-in Skill

create-skill is a built-in Skill that guides you through creating a compliant SKILL.md file interactively. To use it:

/create-skill <Skill description, for example: Convert Word documents to PDF>

Use /create-skill to generate an initial template, then refine it as needed.

Use case: Quickly create custom Skills without learning the file format in detail.

Usage

  1. Automatic trigger: Describe your request. The model determines whether to invoke an appropriate Skill:

Analyze errors in this log file

The model automatically invokes the log-analyzer Skill.

  1. Manual trigger: Enter /skill-name to invoke a specific Skill:

/log-analyzer

Use cases

When to use Skills

  • Complex professional tasks: Workflows that require domain knowledge, such as code review, PDF processing, or API design.

  • Standardized processes: Fixed-step tasks such as commit conventions or deployment workflows.

  • Team knowledge sharing: Packaging best practices for team-wide use.

  • Repetitive work: Frequently performed tasks that require expert guidance.

Examples

Log analysis

Create a log-analyzer Skill that activates when you enter a prompt such as "Analyze this log." It detects errors, performance issues, and abnormal patterns.

API reference generation

Create an api-doc-generator Skill that identifies API endpoints and generates standard documentation and OpenAPI specifications.

Code review

Create a code-reviewer Skill that reviews code against team standards, checking for potential issues and best-practice adherence.