Your actual agent toolkit.
See which built-in tools you rely on, success patterns, MCP usage, skill activation, and orchestration depth.
GitHub 1,204Playwright 891Supabase 744Context7 308
Claude Code integration is live
What's your Agentprint?
Turn the way you build with AI agents into a public, credible profile - tools mastered, context handled, work shipped. No prompts, code, or commands collected.
metadata onlyopen methodologydisconnect anytime
Nova Builder@nova · agent builder
28d liveagent time46hmore than a work week
tool success94%3,842 calls completed
context reuse81%prompt cache leverage
agent activitylast 28 days
agent stackshare of agent use
Claude Code 52%Codex 31%Kiro 17%
GitHub 1,204Playwright 891Supabase 744Context7 308
01Connect once
GitHub sign-in, then one terminal command. No daemon running in the background.
02Build normally
Claude Code exports privacy-filtered OpenTelemetry in the background.
03Share the proof
Your profile turns agent habits into signals other builders can understand.
signal, not vanity
A profile that shows how you build.
Token totals alone are noise. Agentprint makes the operating pattern legible.
Long sessions, handled.
Compaction, cache leverage, and active time reveal whether context is being managed - not merely consumed.
81%cache leverage
Consistency you can see.
Active days and streaks show a real practice, without exposing what you worked on.
17day build streak
Private by construction
The interesting metrics. None of the intimate data.
We reject prompts, responses, tool inputs, outputs, commands, file paths, identities, and custom names before persistence.
prompts blockedsource code blockedcommands blockedusage metadata allowed
Make it your Agentprint.
Start with Claude Code. Connect in under two minutes.
Continue with GitHub