Turbocharge Claude Code, Cursor, Codex, Gemini & every coding agent: faster, cheaper, with contextual understanding specific to your codebase.
TypeScript ai-agentsanthropicclaude-codeclicode-graphcodexcontext-engineeringcursordeveloper-toolsgeminiknowledge-graphllmmcpmcp-serveropen-sourceopenaitree-sitter
docs: hero table = 162-run efficiency metrics (46% tool calls, 42% tokens, 60% time); attribute peak vs table
Shrish Dwivediaab8a83e00f0
diff --git a/README.md b/README.mdindex ee2f10a0..30e0cf8d 100644--- a/README.md+++ b/README.md@@ -17,13 +17,14 @@### Up to **4× cheaper** and **3× faster**, with better or no loss of correctness.-| Averaged across real tasks | Standard Claude Code | With graft |-|---|---|---|-| Cost | baseline | **~29% less** |-| Wall-clock | baseline | **~21% less** |-| Correctness | baseline | **matched** (reproduced the merged PRs) |+| vs. a standard session, no Graft | With Graft |+|---|---|+| Tool calls | **46% fewer** |+| Tokens | **42% fewer** |+| Time | **60% less** |+| Correctness | **equal** |-<sub>Measured on real tasks (real questions + real merged PRs) in popular open-source repos: PocketBase, ollama, Excalidraw. Same agent, same file tools; only the context differs. The "up to 4×" figures are the biggest single-task wins; the table is the average. [Full per-repo numbers ↓](#tested-on-your-popular-repos)</sub>+<sub>The table is a 162-run controlled benchmark (same agent, same file tools, only the context differs). The "up to 4× cheaper / 3× faster" figures are the biggest single-task wins from a separate real-repo sweep (PocketBase, ollama, Excalidraw). [Full method & per-repo numbers ↓](#tested-on-your-popular-repos)</sub></div>