ck

Enable semantic code search across local codebases via the ck MCP server.

6|2|Updated Jul 8, 2012
One-click install
npx skills add https://github.com/athal7/dotfiles --skill ck-athal7
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ck
Source: https://github.com/athal7/dotfiles/tree/main/skills/ck
Command: npx skills add https://github.com/athal7/dotfiles --skill ck-athal7

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Semantic code search across a codebase using the ck MCP server to find meaning-based results and reduce time spent locating implementations.

Core Features & Use Cases

  • Concept discovery — locate whether something exists or where the logic resides.
  • Understanding unfamiliar code — read the region identified by meaning rather than tokens.
  • Finding existing implementations — avoid duplication by locating existing functions and their locations referenced by semantic search.
  • Freshness and performance — lazy indexing per path and roughly 250ms warm query latency.

Quick Start

Run a semantic search query against your codebase to locate relevant implementations.

Frequently Asked Questions about ck

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

FAQPage Schema
How does semantic code search work for finding implementations in a local codebase?

Semantic code search finds implementations by querying meaning rather than exact tokens. It re-embeds paths on the first query, then uses a shared daemon to deliver roughly 250ms warm query latency for fast code discovery.

How do I search for existing code implementations to avoid duplication across worktrees?

To search for existing implementations and avoid duplication, run a semantic query against your local codebase. The tool reindexes on demand and applies fail-open behavior, returning meaning-based results across sessions and worktrees.

What's the best way to locate unfamiliar code logic without knowing exact function names?

The best way to locate unfamiliar code without exact names is semantic code search, which matches conceptual meaning. It identifies relevant code regions by intent, enabling concept discovery even when tokens or function names are unknown.

Does semantic code search require manual reindexing when the local codebase changes?

Semantic code search does not require manual reindexing. It uses lazy indexing per path, reindexing on demand with path re-embedding on the first query, and relies on a single shared daemon to maintain freshness automatically.

What happens when a semantic code search query is slow or the indexing daemon fails?

When a semantic search query is slow, the tool employs fail-open behavior, meaning it returns results without blocking the workflow. This design ensures code discovery and understanding tasks proceed even during latency spikes.