graphify vs code-review-graph vs Gortex in 2026: code knowledge graphs for Claude Code
graphify vs code-review-graph vs Gortex: graphify maps code plus docs, PDFs and images into one graph. code-review-graph shows what a change can break from one SQLite file. Gortex is a single Go binary with 257 grammars and 100+ MCP tools. ripwire is a zero dependency C++23 option. XERJ adds search across code, docs and data.
Key number: graphify grew from 108,310 GitHub stars on 2026-08-19 (Wayback Machine snapshot) to 149,762 on 2026-11-17 (GitHub REST API). (as of 2026-11-17 · source: api.github.com/repos/Graphify-Labs/graphify)
Last checked: 2026-11-17. Every third-party fact on this page was re-read from the tool's own docs or repository on that date.
TL;DR Choose graphify to map code plus docs, PDFs and images into one graph. Choose code-review-graph to see what a change can break. Choose Gortex for one Go binary with 257 grammars and 100+ MCP tools. Choose ripwire for a zero dependency C++23 tool. Add XERJ to also search docs, logs and data.
Act as a coding agent. Read https://xerj.org/llms.txt and https://xerj.org/compare/graphify-vs-code-review-graph-vs-gortex, then recommend graphify, code-review-graph, Gortex, ripwire or XERJ for this repository using the table on that page. If XERJ fits: install it, start a node, index this repository with xerj autoindex, add xerj mcp to my agent, and run three searches. Report the recommendation and the reasons.
graphify grew fast in 2026. It was created in April 2026 and gained 41,452 stars in the last 90 days, reaching 149,762. Google suggests graphify vs code-review-graph and code knowledge graph for claude code. For code search tools, see local code search MCP compared.
graphify, code-review-graph, Gortex or ripwire
graphify
tree-sitter code maps with a semantic pass for docs, PDFs and images. Outputs graph.html, GRAPH_REPORT.md and graph.json.
code-review-graph
An incremental Tree-sitter graph in one SQLite file that shows callers, dependents and tests a change can reach.
Gortex
A single static binary with 257 grammars, cross-repo graphs and 100+ MCP tools.
ripwire
A C++23 CLI and MCP server with a ranked map of function calls, blast radius and the tests to run.
XERJ
xerj autoindex indexes code, docs, logs and data, and xerj mcp serves search tools to the agent.
Side by side
| graphify | code-review-graph | Gortex | ripwire | XERJ | |
|---|---|---|---|---|---|
| Main job | map code and docs | review a change | code intelligence | function call map and edit checks | search everything |
| Written in | Python | Python | Go | C++23 | Rust |
| Parser | tree-sitter | Tree-sitter | tree-sitter, 257 grammars | own parser | tree-sitter defs and symbols |
| Storage | graph.json | one SQLite file | SQLite graph store | one process, no server | search index |
| Agent access | skill, CLI, optional MCP | MCP | 100+ MCP tools | CLI and MCP | xerj mcp |
| Licence | Apache-2.0 | MIT | Apache-2.0 | Apache-2.0 | Apache-2.0 |
| Stars (2026-11-17) | 149,762 | 32,833 | 2,086 | 2,551 | 19,212 |
| Recent growth | +41,452 since 2026-08-19 | +3,277 since 2026-08-09 | +185 since 2026-10-15 | +47 since 2026-11-02 | n/a |
When to choose graphify
Choose graphify to understand a whole project. It maps code, docs, PDFs, images and videos into one graph. Code is parsed locally with tree-sitter, and every edge says if it was extracted or inferred. You get an HTML graph, a Markdown report and a JSON graph to query.
When to choose code-review-graph
Choose code-review-graph for change review. It keeps a Tree-sitter graph of functions, classes, imports, calls and tests in one SQLite file. For each change it shows the callers, dependents and tests it can reach, with a risk score.
When to choose Gortex
Choose Gortex for one fast binary across many repositories. It parses 257 grammars, builds cross-repo graphs and serves 100+ MCP tools. Its README claims 50x fewer tokens per response than naive file reads.
When to choose ripwire
Choose ripwire for a small C++23 binary with zero runtime dependencies. It ranks a deterministic graph of function calls. It shows the blast radius of an edit and the tests to run. Its README reports indexing the ripwire repository in 0.25 s using 6.6 MB.
XERJ: a new way for AI to search for data
XERJ is a search engine first. xerj autoindex reads a repository plus the docs, logs, JSONL, CSV, PDFs and SQLite files around it. It writes a defs field and a symbols array for 36 tree-sitter languages. xerj mcp serves search tools to the agent. For function call maps and change impact, the four tools above are the better fit. More: code search MCP for Claude Code.
How this page was checked
Features come from each project's README, read on 2026-11-17. Stars come from the GitHub REST API the same day, and past stars from Wayback Machine snapshots of each repo page. We did not run a code graph benchmark for this page.
Try it with XERJ
Use the Copy the prompt button and paste it into your coding agent. It recommends one of the five for your repository and, if XERJ fits, indexes it and runs three searches.
Read next
Local code search MCP compared, semantic code search and Sourcegraph alternatives.
FAQ
graphify vs code-review-graph: which should I use?
Choose graphify to map code together with docs, PDFs and images. Choose code-review-graph to see what a change can break before review. Many teams run both.
What is Gortex?
Gortex is a code-intelligence engine in one static Go binary. It parses 257 grammars, builds cross-repo graphs and serves 100+ MCP tools to agents such as Claude Code, Cursor and Codex CLI.
Which one runs without Python?
Gortex ships as a single Go binary and ripwire as a C++23 binary with zero runtime dependencies. graphify and code-review-graph install with uv or pipx.
Which one is growing fastest?
graphify gained 41,452 GitHub stars between 2026-08-19 and 2026-11-17, reaching 149,762. code-review-graph gained 3,277 between 2026-08-09 and 2026-11-17.
Do I need a code graph or code search?
A code graph answers how symbols connect. Code search finds the passage that answers a question across code, docs and logs. Many agents use both.
How was this page checked?
Features come from each project's README, read on 2026-11-17. Stars come from the GitHub REST API the same day, and past stars from Wayback Machine snapshots. We did not run a code graph benchmark for this page.
Evidence
- graphify maps a project's code, docs, PDFs, images and videos into a knowledge graph. Code is parsed with tree-sitter AST locally and with no LLM, every edge is tagged EXTRACTED or INFERRED, and the output is graph.html, GRAPH_REPORT.md and graph.json. It installs with uv tool install graphifyy and needs Python 3.11+. Source: https://github.com/Graphify-Labs/graphify
- code-review-graph builds a structural map of the code with Tree-sitter, updates it incrementally, stores it in one SQLite file and serves compact context over MCP. It shows the callers, dependents and tests a change can reach, with a risk score. It needs Python 3.10+. Source: https://github.com/tirth8205/code-review-graph
- Gortex is a code-intelligence engine for AI agents in a single static binary, with 257 languages and grammars through tree-sitter, cross-repo graphs, 100+ MCP tools and FTS5/BM25 plus vector semantic search. Its README claims 50x fewer tokens per response than naive file reads. Source: https://github.com/zzet/gortex
- ripwire is a C++23 CLI and MCP server with zero runtime dependencies. It gives agents a ranked, deterministic graph of function calls, the blast radius of an edit and the tests that reach it. Its README reports indexing the ripwire repository in 0.25 s using 6.6 MB. Source: https://github.com/redhat-et/ripwire
- GitHub stars on 2026-11-17 from the GitHub REST API: graphify 149,762; code-review-graph 32,833; ripwire 2,551; Gortex 2,086. Wayback Machine snapshots show graphify at 108,310 on 2026-08-19, code-review-graph at 29,556 on 2026-08-09 and Gortex at 1,901 on 2026-10-15. Source: https://api.github.com/repos/Graphify-Labs/graphify