Lin-A1

8 mods across 1 repository, 7 stars between them.

deepsearch-service

01

Lin-A1/skills-agent

Skill Claude CodeCodex

A deep-search and reasoning service based on large language models. It breaks complex questions into smaller searches, evaluates and verifies information, and produces a structured analysis report.

7 7mo ago A 57 tokens

embedding-service

02

Lin-A1/skills-agent

Skill Claude CodeCodex

A text-embedding service that converts natural-language text into numerical vectors. These vectors help software compare meaning rather than only matching exact words.

7 7mo ago A 45 tokens

ocr-service

03

Lin-A1/skills-agent

Skill Claude CodeCodex

Optical character recognition (OCR) for finding and extracting written text from images in multiple languages and formats. It also reports each text area's position and a confidence score.

7 7mo ago A 53 tokens

rag-service

04

Lin-A1/skills-agent

Skill Claude CodeCodex

A retrieval-augmented generation service, meaning it finds relevant passages from stored documents before they are used to produce an answer. It stores documents in Milvus, a vector database, and reorders search results with a ranking model.

7 7mo ago A 54 tokens

rerank-service

05

Lin-A1/skills-agent

Skill Claude CodeCodex

A service for reordering search results by relevance. It can return complete ranked results or just the indexes of the most relevant documents.

7 7mo ago A 55 tokens

sandbox-service

06

Lin-A1/skills-agent

Skill Claude CodeCodex

A Docker-based service for running Python, Shell, or Bash code in an isolated environment with time and resource limits.

7 7mo ago A 44 tokens

search-guide

07

Lin-A1/skills-agent

Skill Claude CodeCodex

A decision guide for choosing when an agent should use a search tool and when it should rely on its existing knowledge.

7 7mo ago A 48 tokens

websearch-service

08

Lin-A1/skills-agent

Skill Claude CodeCodex

A real-time web-search service built around SearXNG and visual-language-model analysis. It searches current information, extracts webpage content, and uses two levels of caching to reuse recent results.

7 7mo ago A 53 tokens