Changelog

All notable changes to Gweta will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

0.1.0 - 2025-02-06

Added

Core

• Chunk dataclass for universal chunk representation
• QualityReport and QualityIssue types
• GwetaSettings configuration management
• Structured logging with get_logger()

Validation (4 Layers)

• Layer 1: Extraction Quality
• ExtractionValidator for raw text validation
• Gibberish detection
• Encoding validation

• Language detection

• Layer 2: Chunk Quality

• ChunkValidator for chunk validation
• Information density scoring
• Coherence checking

• DuplicateDetector using MinHash LSH

• Layer 3: Domain Rules

• DomainRuleEngine with YAML configuration
• Regex pattern matching
• Numerical range validation
• Required field validation

• Known fact cross-referencing

• Layer 4: KB Health

• HealthChecker for ongoing monitoring
• Staleness detection
• Duplicate identification
• GoldenDatasetRunner for retrieval testing
• JUnit XML and JSON export

Acquisition

• GwetaCrawler wrapping Crawl4AI
• PDFExtractor with table extraction
• DatabaseSource with SQL safety
• APIClient for REST endpoints
• Sitemap and RSS fetchers

Ingestion

• RecursiveChunker for text splitting
• ChromaStore adapter
• QdrantStore adapter
• PineconeStore adapter
• WeaviateStore adapter
• BaseStore abstract interface

MCP Server

• FastMCP-based server
• 10 MCP tools for AI agent integration
• 4 MCP resources
• 3 MCP prompts
• stdio and HTTP transports

CLI

• gweta validate command
• gweta crawl command
• gweta ingest command
• gweta health command
• gweta serve command

Documentation

• Getting started guide
• Architecture overview
• API reference
• Example scripts

---

Version History

Version	Date	Description
0.1.0	2025-02-06	Initial release