Launching today
Graphicle
A Rust port of Graphify, plus memory for your coding agents
1 follower
A Rust port of Graphify, plus memory for your coding agents
1 follower
Graphicle is a Rust port of Graphify's open-source code-graph engine, 3 to 22.7x faster than Graphify 0.9.91 in all 39 of our measurements, plus project memory: docs, decisions and past agent sessions, cited to file and line over MCP. Local, Apache-2.0.
Hi Product Hunt! I'm Vikram, and I built Graphicle.
Why:
A coding agent starts every session without a map of a big repository. Graphify's code graph gives it one, but on kubernetes it took 25 minutes to build in our runs. So I ported its engine to Rust, which builds the same repository's graph in 1 min 8 s, and added memory, so the agent also gets your docs, decisions and past sessions as cited excerpts.
Memory, compared:
In early pre-release evaluations on GPT models, under Mem0's own LoCoMo protocol (its prompts, GPT-5 as reader and judge), Graphicle scored 93.5% against 91.6% in Mem0's published result files. In a pre-release comparison of seven memory systems on one open model (development sets), it matched Hindsight 0.10.2 and beat Mem0 OSS 2.2.1, while making about 12 model calls per conversation to ingest it, against about 160 for Hindsight and 332 for Mem0. Every number and its labels are in the repo.
Who it's for:
Anyone using Claude Code, Codex, Cursor or another MCP client on a large codebase, and Graphify users who want the same graph faster, with memory built in.
New in 0.1.8:
the median question on kubernetes went from 2.99 s to 1.34 s (our server, 22 questions);
the video races the current Graphify 0.9.95 on scikit-learn: 3.6 s against 30.2 s, one run each, 24 cores each.
Try it:
Which repository is slow or gets poor context for you?
GitHub · free and open source (Apache-2.0)