Skip to content
SolidRusT.ai

Changelog

All notable changes to the SolidRusT AI API are documented here.

This changelog follows Keep a Changelog format and the API adheres to Semantic Versioning.


Features and changes in development, not yet deployed to production.


[2026-06-20] - Full API Documentation Alignment

Section titled “[2026-06-20] - Full API Documentation Alignment”
  • New Graph Traversal endpoints documented: /data/v1/query/graph, /graph/search, /graph/entity/{name}, /graph/documents
  • New Document Ingestion endpoints documented with idempotency notes
  • New API Reference pages for Semantic Search, Keyword Search, Hybrid Search, Knowledge Graph, Graph Traversal, Ingestion, Agent Chat, and Agent Tools
  • New Management endpoints documented: /data/v1/sources, /data/v1/stats
  • New Agent memory tools documented: MemorySearchTool, MemoryStoreTool, MemoryStatsTool
  • New ROADMAP.md and Docs Sync Process in CLAUDE.md for tracking future API → docs gaps
  • Changed Hybrid search schema updated: added use_entity_extraction, game, use_graph, graph_depth fields
  • Changed Agent tools count updated from 3 to 6 tools
  • Changed OpenAPI spec expanded from 9 to 22 endpoints (v2.0)
  • Changed API Overview page updated with complete endpoint listing
  • Changed Sidebar reorganized with 8 new API Reference entries
  • Fixed Hybrid search response schema corrected to include vector_count, graph_count, merged_count, entities_found
  • Fixed Agent tool descriptions updated to match actual registered tools
  • Fixed Embeddings dimension documented as 1024 (was generically described)

  • New Privacy-focused analytics on documentation site
  • New Comprehensive SEO metadata for all public sites
  • New Sitemap generation for improved search indexing
  • Improved API documentation with interactive OpenAPI explorer

  • New API Gateway at api.solidrust.ai - Public endpoint for all API calls
  • New Documentation Site at docs.solidrust.ai - Comprehensive API documentation
  • New PAM Platform - API key management and billing via console.solidrust.ai
  • /v1/chat/completions - OpenAI-compatible chat completions
  • /v1/embeddings - Vector embeddings using bge-m3 model
  • /v1/models - List available models
  • /data/v1/query/semantic - Semantic search (RAG)
  • /data/v1/query/keyword - Keyword search (RAG)
  • /data/v1/query/hybrid - Hybrid search (RAG)
  • /v1/agent/chat - Tool-enabled agent conversations

  • Initial vLLM inference deployment with Qwen3-4B model
  • LiteLLM proxy for OpenAI compatibility
  • Embeddings service with bge-m3 model
  • Kubernetes cluster with 12 nodes (including 5 GPU nodes)
  • Automatic failover to Claude Haiku when vLLM is unavailable

BadgeMeaning
NewNew feature or endpoint
ChangedChanges to existing functionality
DeprecatedFeature scheduled for removal
RemovedRemoved feature
FixedBug fix
SecuritySecurity-related change


Found a bug or have a feature request? Open an issue on GitHub.