Don't Let Reasoning
Die in the Tab.
Grov automatically captures the context from your private AI sessions and syncs it to a shared team memory.
~500 tokens injected • No file exploration • Verified team knowledge
The problem
Every session, Claude re-explores your codebase.
It reads the same files. Rediscovers the same patterns.
You burn tokens on redundant exploration.
How Grov saves tokens
Semantic search finds relevant memories, shows lightweight previews, expands only what's needed.
Measured on real Claude Code sessions. Results may vary based on codebase complexity.
How it works
Install once
Three commands. Takes 30 seconds.
$ npm install -g grov
$ grov init
$ grov proxy Use Claude normally
Keep grov proxy running in a terminal. It captures reasoning from your sessions through a local proxy. No changes to your workflow.
Context compounds
Throughout your session, grov injects relevant context from past sessions. Claude skips exploration and reads files directly. Your AI gets smarter over time.
Stay on track
Grov monitors what Claude does and corrects if it drifts from your goal. It scores alignment and injects corrections automatically.
What gets captured
Not what changed. Why it changed.
{
"task": "Fix auth logout bug",
"goal": "Prevent random user logouts",
"files_touched": [
"src/auth/session.ts",
"src/middleware/token.ts"
],
"reasoning_trace": [
"Investigated token refresh logic",
"Found refresh window was too short",
"Extended from 5min to 15min"
],
"status": "complete",
"tags": ["auth", "session", "token"]
}The "why" behind each decision, not just the code diff
Files touched, patterns used, constraints discovered
Auto-generated for quick retrieval and team search
Get started
Three commands. That's it.
Requires Node.js 18+ and Claude Code v2.0+
