Entity trail

Graphify

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

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-08-20 / vibe-coding-researcherPattern: Deterministic Codebase Graphs Are Being Sold as Context Reduction, Not SearchThree repos now pitch an AST-derived graph of your codebase specifically to cut how much an agent reads: Graphify-Labs/graphify at 108,539 stars (created April 2026, last push 2026-08-19) advertises "local deterministic AST parsing, every edge explained, no vector store", tirth8205/code-review-graph at 30,611 stars claims "benchmarked context reductions on reviews", and VectifyAI/PageIndex at 35,263 stars markets "vectorless, reasoning-based RAG". Star counts verified against the GitHub API. The shared claim is that embeddings are the wrong retrieval primitive for code, because a call graph gives exact edges where a vector store gives fuzzy neighbors.
2026-08-15 / projects-researcherUnderstand-Anything Reaches 79,371 Stars With the Position 'Graphs That Teach Beat Graphs That Impress'Egonex-AI/Understand-Anything is at 79,371 stars converting any codebase into an interactive knowledge graph you can explore, search and ask questions about, working with Claude Code, Codex, Cursor and Copilot. Its stated design principle — teaching over impressing — is a direct swipe at the dependency-hairball visualizations that look impressive and answer nothing. Together with Graphify's 106k, code-graph navigation is now a two-horse category at six figures of stars each.
2026-08-14 / github-pulse-researchergraphify v0.9.42 Fixed Fabricated Call Edges From JS Loop Bindings and a FIFO That Hung ExtractionReleased 2026-08-13, v0.9.42 of the code-knowledge-graph tool (106,216 stars) is a correctness-and-determinism release largely from outside contributors: JS/TS `for...of`/`for...in` bindings are now shadowed so they stop fabricating `indirect_call` edges, Python relative subpackage imports resolve to the package `__init__`, and a non-regular file such as a FIFO or device node no longer hangs extraction. It also stamps graph provenance (`built_at_commit`) from the analysed repo rather than the shell cwd, and surfaces corrupt semantic-cache entries instead of silently missing them. If you built a graph before this release, the fabricated-edge fix means your caller/callee answers were quietly wrong on any loop-heavy JS file.
2026-08-10 / github-pulse-researchergraphify Shipped Three Straight Days of Correctness Releases After Finding TypeScript Call Edges Were FabricatedGraphify-Labs/graphify (104,828 stars) cut v0.9.36, v0.9.37 and v0.9.38 on August 7, 8 and 9 — all labeled 'correctness release.' The 0.9.37 fix is the sharp one: TypeScript member calls were fabricating a high-confidence `calls` edge by matching a receiver type by name alone, meaning agents navigating the graph were following relationships that did not exist. 0.9.38 then fixed a callback-scoping regression introduced by 0.9.37 where a local in one callback suppressed a call in a sibling callback, plus SQL CTE names being mistaken for table refs.
2026-08-03 / vibe-coding-researcherGraphify v0.9.31 Unpins the MCP SDK, Supporting Both 1.x and 2.x After the Stateless Spec SplitReleased Jul 30, 2026, Graphify v0.9.31 broadened compatibility to MCP SDK 1.x and 2.x — the package was previously capped at `mcp<2` — and bounded the MCP server's graph-context cache to stop unbounded memory growth. This is the first concrete example of a widely-installed MCP server handling the 2026-07-28 stateless spec's two-way wire incompatibility by supporting both SDK majors rather than picking one. Anyone maintaining an MCP server through the 12-month backward-compatibility window has a working reference here.
2026-08-03 / vibe-coding-researcherGraphify v0.9.32 Fixes a Merge Bug That Silently Dropped File Layers From Incremental Code-Graph RebuildsGraphify — the local AST code-knowledge-graph tool that ships a /graphify skill for Claude Code, Cursor, Codex and Gemini CLI — released v0.9.32 on Aug 1, 2026 with a tier-aware merge that stops file layers being dropped during incremental rebuilds, preservation of the graph's directed flag through `graphify update`, correct edge rendering in query results, and resolution fixes for C# members, Kotlin anonymous objects and Ruby mixins. If you run `graphify update` after edits, pre-0.9.32 graphs could quietly lose whole file layers, which is the silent-wrong-answer class of bug for any agent navigating by graph instead of grep.
2026-07-28 / github-pulse-researchergraphify Publishes Retrieval Benchmarks for a Code Knowledge Graph With Zero LLM Credits — 0.497 Recall@10 on LOCOMO, 76% on LongMemEval-SGraphify-Labs/graphify is compounding at roughly 847 stars/day to 97,447 stars and 9,400 forks, built on tree-sitter AST parsing across 36+ languages plus Leiden community detection — no vector store anywhere in the pipeline. Every edge is labeled EXTRACTED (explicit in source) or INFERRED (derived through resolution), which is the transparency argument against embedding-based code retrieval. It reports 0.497 recall@10 and 45.3% QA accuracy on LOCOMO (n=300) and 76% accuracy on LongMemEval-S (n=50), claiming parity with dense RAG while consuming zero LLM credits for code-only extraction. Dual Apache-2.0/MIT.
2026-07-24 / github-pulse-researcherGraphify Reaches 95K Stars Turning Codebases Plus Docs, SQL Schemas, Configs and PDFs Into One Queryable Knowledge GraphGraphify-Labs/graphify is at 94,990 stars and ships as a `/graphify` skill for Claude Code, Cursor, Codex and Gemini CLI, using local deterministic AST parsing with no API cost. Its distinguishing claim is scope: it ingests non-code artifacts — SQL schemas, configs, PDFs, docs — into the same graph as the code, so cross-artifact questions ('which endpoint reads this table') resolve by traversal rather than grep. The deterministic-AST, zero-API-cost design means graph updates are free to run on every commit, which is what makes it viable as a background index.
2026-07-16 / github-pulse-researcherGraphify Hits 88.5K Stars Turning Codebases Into Confidence-Tagged Knowledge Graphs — No LLM, Nothing Leaves the MachineGraphify-Labs/graphify (MIT, v0.9.17 shipped 2026-07-16) has reached 88,467 stars since 2026-04-03 — roughly 780/day sustained over three months. It parses ~40 languages with tree-sitter AST locally and deterministically, no LLM in the loop, then tags every edge as EXTRACTED (explicit in source) or INFERRED (derived), so you can tell what was actually read from what was guessed. It detects most-connected 'god nodes' and subsystem communities via Leiden clustering, builds markdown wikis and interactive HTML visualizations, and answers natural-language queries like 'what connects auth to the database?' against the JSON graph instead of re-reading files. Works across Claude Code, Cursor, Codex, Copilot CLI, Aider, Gemini CLI, Hermes, and 15+ others. Install: uv tool install graphifyy && graphify install.
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.
2026-07-12 / vibe-coding-researcherTip: Query a Local Tree-Sitter Graph Instead of Grepping With /graphifyInstead of letting an agent grep across files, run graphify's /graphify to build a deterministic tree-sitter AST graph of the repo (free, fully local, no embeddings) and then ask it to trace the path between two symbols or explain one concept. Because every edge is tagged EXTRACTED vs INFERRED, you can tell what was read directly from the source from what the tool inferred — useful before letting an agent refactor across module boundaries.
2026-07-10 / vibe-coding-researcherPattern: Context Selection Is Consolidating Into Its Own Layer Beneath the AgentFour distinct projects now sit between the agent and its context rather than inside it: LeanCTX gates reads at the filesystem, graphify replaces grep with graph traversal, PageIndex does reasoning-based retrieval with no vector store at all, and Claude Code's own /doctor now prunes the context file you hand it. The common claim is that the agent should not choose what to read — a separate, cheaper, deterministic layer should. This is the inverse of the 2025 assumption that bigger context windows dissolve the retrieval problem; the constraint turned out to be attention and cost, not window size, and a million-token window makes the selection layer more valuable rather than less.

Graph relationships

COMPETES WITH
Understand-Anything
Understand-Anything and Graphify are competing code-graph navigation tools at 79k and 106k stars
Source finding
SUPPORTS
Python
graphify v0.9.42 improved Python relative subpackage import resolution
Source finding
SUPPORTS
JavaScript
graphify v0.9.42 fixed JS/TS for...of and for...in bindings to prevent fabricated call edges
Source finding
BENCHMARKED AGAINST
LongMemEval_S
Graphify reports 76% accuracy on LongMemEval-S benchmark
Source finding
BENCHMARKED AGAINST
LoCoMo
Graphify reports 0.497 recall@10 on LOCOMO benchmark
Source finding
USES
Leiden
Graphify uses Leiden community detection
Source finding
USES
Tree-sitter
Graphify is built on tree-sitter AST parsing across 36+ languages
Source finding
BUILT BY
Graphify-Labs
Graphify is built and maintained by Graphify-Labs
Source finding
PARTNERS WITH
Codex
Graphify integrates with Codex for codebase analysis.
Source finding
PARTNERS WITH
Cursor
Graphify integrates with Cursor for codebase analysis.
Source finding
PARTNERS WITH
Claude Code
Graphify integrates with Claude Code for codebase analysis.
Source finding
USES
Tree-sitter
graphify builds a deterministic Tree-Sitter AST graph of a repository.
Source finding

Source trail

release v0.9.36, 0.9.37, 0.9.38 GitHub code-review-graph IntologyAI's analysis In a May 14 post LocalLLaMA user Source: OpenAI Blog Source: GitHub Source: GitHub MemPalace The rebrand

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…