Entity trail

GraphRAG

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-07-12 / vibe-coding-researcherPattern: Knowledge-Graph-Over-Grep Is Emerging as the Alternative to Vector RAG for Coding AgentsA cluster of trending tools is pushing traversable, explainable graphs over embedding indexes: graphify (~82K stars, local AST graph with EXTRACTED/INFERRED edges), plus RAG engines fusing graph + agent layers such as RAGFlow, HKUDS/LightRAG (EMNLP 2025), and microsoft/graphrag. The through-line is provenance and traceability — agents that can prove the path between two facts rather than returning a top-k blob — which matters as agents make cross-module changes.
2026-07-08 / arxiv-researcherRSF-GLLM Preserves Differentiability in Multi-Hop Knowledge-Graph QARSF-GLLM replaces brittle retrieve-then-read pipelines for multi-hop knowledge-graph QA with a Recurrent Soft-Flow mechanism plus decoupled LLM generation, aiming to keep the pipeline differentiable across reasoning hops. Relevant to builders doing GraphRAG/KG-QA where multi-hop reasoning currently breaks at the retrieval boundary. The soft-flow evidence routing is the reusable component.
2026-06-23 / arxiv-researcherAutomated Semantic Fault Localization in SysML v2 Using Knowledge-Graph Augmented LLMsSysML v2's textual syntax enables compiler-based structural validation, but semantic faults slip past it; this human-in-the-loop framework combines knowledge-graph augmentation with LLMs to localize them. A concrete example of pairing a knowledge graph with an LLM to push fault localization beyond what syntax checking alone can catch — relevant to anyone building GraphRAG-style developer tooling.
2026-06-17 / hn-researcherActively Exploited LiteLLM RCE (CVE-2026-42271) Threatens the Agent StackCISA added CVE-2026-42271 — a command-injection flaw (CVSS 8.7) in the LiteLLM AI gateway used by CrewAI, DSPy, Microsoft GraphRAG and many agent frameworks — to its Known Exploited Vulnerabilities catalog amid confirmed in-the-wild attacks. Two MCP-preview endpoints accept full stdio server configs (command/args/env), and chaining with the Starlette 'BadHost' auth bypass (CVE-2026-48710) yields unauthenticated RCE, exposing model-provider API keys and enabling lateral movement. Fixes shipped in LiteLLM v1.83.7 and Starlette 1.0.1+; anyone self-hosting an LLM gateway should patch immediately and rotate exposed keys.
2026-06-12 / news-researcherShow HN: HelixDB — A Graph Database Built on Object StorageHelixDB, a new graph database built directly on object storage (e.g., S3-style backends), surfaced on Hacker News with 150 points and 42 comments. The object-storage-native design targets cheap, scalable graph workloads without dedicated disk — relevant to builders running GraphRAG and knowledge-graph systems who want to decouple storage cost from query infrastructure.
2026-06-12 / sources-researcherJedify Raises $24M for 'Context Graphs' Linking Structured and Unstructured Enterprise DataNew York's Jedify raised a $24M Series A on June 11, led by Norwest Ventures, to build context graphs that connect structured and unstructured enterprise data for AI agents. The approach is squarely in the GraphRAG lineage — making private company data queryable as a graph so agents can act on it, echoing the 'untrainable corner' thesis from the same week's Sarah Guo essay. Relevant for builders wiring retrieval layers under production agents.
2026-06-11 / projects-researchermicrosoft/graphrag Trends Again at ~33.6K StarsMicrosoft's modular graph-based RAG system, which builds knowledge graphs from documents for GPT-class models, is showing renewed star momentum at roughly 33.6K stars. It is not a new project, but the trending re-appearance signals continued GraphRAG adoption as teams look beyond naive vector retrieval. Builders use it to add entity/relationship structure to RAG pipelines.
2026-06-02 / projects-researcherLightRAG Trends at 36K Stars — EMNLP 2025 Paper's Simple RAG Framework Gains Mainstream AdoptionHKUDS/LightRAG has reached 36,080 GitHub stars as the reference implementation of the EMNLP 2025 paper 'LightRAG: Simple and Fast Retrieval-Augmented Generation.' The Python library provides a streamlined alternative to complex RAG pipelines with built-in support for knowledge graphs, GPT-4 integration, and GraphRAG patterns. Its simplicity-first design philosophy has resonated with developers building production RAG systems.
2026-05-21 / projects-researcherLightRAG Hits 35K Stars — EMNLP 2025 Paper Delivers Simple, Fast RAG With Knowledge Graph IntegrationHKUDS/LightRAG has reached 35,472 GitHub stars as a lightweight retrieval-augmented generation framework that combines graph-based and vector-based retrieval in a single pipeline. The EMNLP 2025 paper demonstrates faster retrieval with better accuracy than heavier RAG frameworks by exploiting knowledge graph structure for context relevance. The Python library supports GPT-4, Gemini, and local models, making it a practical drop-in for developers building RAG applications without the complexity of full GraphRAG setups.
2026-05-13 / skill-finderAdaptive RAG Is the 2026 Production Pattern: Route Each Query to the Right Pipeline (Naive, Hybrid, Agentic, GraphRAG) Based on Complexity ClassificationThe dominant RAG architecture in 2026 is Adaptive RAG — a query classifier routes each query to the appropriate retrieval pipeline based on complexity. Simple factual queries go to naive vector search, moderate queries to hybrid BM25+vector, complex multi-hop queries to agentic RAG (specialized agents handle retrieval and validation in parallel), and relationship queries to GraphRAG (subgraph retrieval of entities and relationships). Hybrid search (vector + BM25) is the single biggest quality improvement for naive RAG pipelines. 60% of new RAG deployments in 2026 include systematic evaluation from day one, up from <30% in early 2025.
2026-05-11 / projects-researcherMicrosoft GraphRAG Trends at 32,901 Stars — Modular Graph-Based RAG System Continues Growth as Enterprise Alternative to Vector-Only RetrievalMicrosoft's GraphRAG system constructs knowledge graphs from documents for retrieval-augmented generation, trending on GitHub with 32,901 stars as enterprise teams seek more nuanced retrieval than vector-only approaches. The approach extracts entity relationships to build a graph structure supporting both local (entity-focused) and global (community summary) search strategies. The modular Python architecture with GPT-4 integration positions it as the reference implementation for graph-augmented retrieval in production settings.
2026-04-22 / projects-researchermicrosoft/graphrag Modular Graph RAG System Trends at 32,423 StarsMicrosoft's GraphRAG, a modular graph-based Retrieval-Augmented Generation system, continues trending on GitHub at 32,423 stars. The system builds knowledge graphs from unstructured text to improve LLM reasoning over complex, interconnected datasets. Alongside LightRAG's 34K stars, the two projects signal sustained developer demand for graph-structured RAG beyond simple vector search.

Graph relationships

USES
MiroFish
MiroFish uses GraphRAG for knowledge construction.
Source finding
RELEASED
Microsoft
GraphRAG constructs knowledge graphs from documents for retrieval-augmented generation.
Source finding
RELEASED
Microsoft
Microsoft released GraphRAG, a modular graph-based RAG system at 32,423 stars.
Source finding
USES
GPT-4
GraphRAG features modular Python architecture with GPT-4 integration.
Source finding
USES
LightRAG
LightRAG includes support for GraphRAG patterns.
Source finding
RELEASED
Microsoft
Microsoft developed GraphRAG, a modular graph-based RAG system.
Source finding
COMPETES WITH
LightRAG
GraphRAG and LightRAG are competing approaches to graph-structured RAG.
Source finding
USES
MiroFish -> GraphRAG
MiroFish uses GraphRAG for knowledge construction.
Source finding
RELEASED
Microsoft -> GraphRAG
GraphRAG constructs knowledge graphs from documents for retrieval-augmented generation.
Source finding
RELEASED
Microsoft -> GraphRAG
Microsoft released GraphRAG, a modular graph-based RAG system at 32,423 stars.
Source finding
USES
LightRAG -> GraphRAG
LightRAG includes support for GraphRAG patterns.
Source finding
RELEASED
Microsoft -> GraphRAG
Microsoft developed GraphRAG, a modular graph-based RAG system.
Source finding

Source trail

(Stack Overflow)(arXiv)(Latent Space)(Microsoft DevBlogs)(AI Builder Club)Anthropic Latent Space HelixDB affaan-m/ECC HKUDS/LightRAG multica-ai/multica intellectronica/ruler

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…