semble

Locate implementations, symbols, and related code via natural language or identifier queries.

5|Updated Feb 8, 2026
One-click install
npx skills add https://github.com/tiancaiamao/ai --skill semble-tiancaiamao
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: semble
Source: https://github.com/tiancaiamao/ai/tree/main/skills/semble
Command: npx skills add https://github.com/tiancaiamao/ai --skill semble-tiancaiamao

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you quickly locate implementations, symbols, and related code in unfamiliar repositories without relying on slow manual grep-and-read workflows.

Core Features & Use Cases

  • Natural Language Code Search: Ask for features, behaviors, or implementation points in plain English and get relevant code matches.
  • Symbol and Related-Code Discovery: Jump from a file or line to similar code paths and connected logic across the repository.
  • Repository Exploration: Use it when onboarding to a new codebase, tracing a feature, or finding where a behavior is implemented.

Quick Start

Use the semble skill to search the repository for the implementation of a feature you want to understand.

Frequently Asked Questions about semble

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I search a codebase using natural language instead of grep?

Natural language code search allows you to query features or behaviors in plain English to locate implementations and symbols, reducing manual grep-and-read work across source files, docs, and config.

Can I find connected code and related symbols across a repository offline?

Yes, offline indexing supports repository exploration by locating implementations, symbols, and related code across source files without requiring continuous network connectivity after model download.

What is the best way to navigate an unfamiliar codebase for feature tracing?

Repository exploration tools help you navigate unfamiliar codebases by using natural language or identifier queries to discover feature implementations and jump between connected logic paths across files.

How to locate specific implementations and symbols in source files efficiently?

Locating implementations and symbols is efficient with natural language or identifier queries that index source files and config locally, returning relevant code matches while reducing manual grep-and-read work.

Does code search with offline indexing work without an internet connection?

Yes, offline-capable code search functions without an internet connection after the model download is complete, utilizing efficient local indexing to deliver relevant code matches for repository exploration.

Why should I use natural language search instead of manual grep for repository exploration?

Natural language search reduces manual grep-and-read work by allowing you to ask for features, behaviors, or implementation points in plain English, returning relevant code matches and connected logic across the repository.