Graphify
Graphify is an open-source tool for constructing knowledge graphs, capable of transforming unstructured data such as documents and text into structured graph form, supporting knowledge extraction, association, and visual management. Its main characteristics include an efficient data processing mechanism and low computational resource consumption, making it applicable to knowledge management, information retrieval, and natural language processing scenarios.
Timeline (last 90 days)
According to the video, Graphify can cut the token consumption of codebase exploration by more than 80%.
Graphify's `graphify extract . --code-only` command parses source code with tree-sitter and generates `graph.json`; this step incurs no API cost and is pure local static analysis.
`graphify cluster-only` additionally produces `GRAPH_REPORT.md` (a text summary for the model) and an interactive `graph.html` (a visualization for humans).
For each question, RepoBuddy first runs `graphify query` to return the relevant subgraph, then sends it along with the report to Gemma 4, and the prompt hard-requires answering only from the graph and giving file paths.
When `graphify query` returns exit code 4 (the question is too broad and no node matches), RepoBuddy falls back to answering from the report alone rather than erroring out.
RepoBuddy's workflow has three steps: shallow-clone the repository, build a knowledge graph with Graphify, and after graph retrieval hand the question to Gemma 4 to answer.
Graphicle is a project that rewrites Graphify in Rust and adds memory capability for coding agents on top of the code graph.
In a real-world test of adding a field to the CRM, the average cost was 1.5 million tokens without Graphify, 1.2 million with Graphify installed but not actively invoked (Claude did not actually use it), and approximately 1.35 million when explicitly required to use it, failing to prove that Graphify saves money
Graphify can convert an open-source CRM project with over 800 files into a graph containing 3,000 interconnected nodes grouped by application area within 5 seconds
All Facts (10)
Graphify图谱引擎相比传统检索方式，节省了高达71倍的Token消耗
60%UnverifiedAccording to the video, Graphify can cut the token consumption of codebase exploration by more than 80%.
50%UnverifiedWhen `graphify query` returns exit code 4 (the question is too broad and no node matches), RepoBuddy falls back to answering from the report alone rather than erroring out.
50%UnverifiedRepoBuddy's workflow has three steps: shallow-clone the repository, build a knowledge graph with Graphify, and after graph retrieval hand the question to Gemma 4 to answer.
50%UnverifiedGraphify's `graphify extract . --code-only` command parses source code with tree-sitter and generates `graph.json`; this step incurs no API cost and is pure local static analysis.
50%Unverified`graphify cluster-only` additionally produces `GRAPH_REPORT.md` (a text summary for the model) and an interactive `graph.html` (a visualization for humans).
50%UnverifiedFor each question, RepoBuddy first runs `graphify query` to return the relevant subgraph, then sends it along with the report to Gemma 4, and the prompt hard-requires answering only from the graph and giving file paths.
50%UnverifiedGraphicle is a project that rewrites Graphify in Rust and adds memory capability for coding agents on top of the code graph.
50%UnverifiedGraphify can convert an open-source CRM project with over 800 files into a graph containing 3,000 interconnected nodes grouped by application area within 5 seconds
50%UnverifiedIn a real-world test of adding a field to the CRM, the average cost was 1.5 million tokens without Graphify, 1.2 million with Graphify installed but not actively invoked (Claude did not actually use it), and approximately 1.35 million when explicitly required to use it, failing to prove that Graphify saves money
50%