Can I use this for vector search implementation?

Yes, it includes guidance on integrating HNSW vector search, including handling dimension mismatches, cosine normalization, and hybrid search paths.

What is the primary purpose of the Index Lifecycle skill?

It provides the mental model and authoritative sequence for managing Searchlite's storage, ensuring developers can correctly modify ingestion, durability, and search logic.

How does this skill handle data durability?

It outlines the commit contract where segment files and manifests are fsync’d, and WAL truncation only occurs after a successful, atomic manifest update.

What is the role of the Manifest in Searchlite?

The Manifest is the authoritative list of segments that define the current index state; it is updated atomically during the commit process.

Where should core indexing changes be made in the codebase?

According to the skill guidance, core indexing, query, and durability logic should be modified within the searchlite-core package to maintain source-of-truth correctness.

Searchlite Index Lifecycle

Name: Searchlite Index Lifecycle
Author: davidkelley

bydavidkelley

0•

データベース管理

Manages and optimizes Searchlite's internal storage pipeline, including indexing, durability, and segment compaction.

The Searchlite Index Lifecycle skill provides a comprehensive guide and operational framework for developers working with the Searchlite embedded search engine. It details the internal mechanics of the indexing pipeline, from Write-Ahead Logs (WAL) and segment builds to manifest management and data compaction. This skill is essential when implementing custom storage backends, optimizing ingestion performance, ensuring data durability, or modifying schema-driven full-text search behavior within the Searchlite ecosystem, providing specific guidance on maintaining atomicity and schema invariants.

主な機能

01Fast-field and schema invariant enforcement

02Internal storage pipeline management (WAL to segment build)

03Detailed manifest and compaction logic guidance

04Vector search (HNSW) and hybrid search integration

05Durability and crash recovery implementation patterns

060 GitHub stars

ユースケース

01Implementing custom storage backends for WASM or FFI environments

02Building or modifying the core indexing and query durability layers in Rust

03Optimizing full-text search performance and aggregation efficiency

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add davidkelley/searchlite index-lifecycle

For use in Claude.ai and ChatGPT

Download Skill