AI Insights

Rūta

@rtuckuteJoined Nov 2026covers Aug 28, 2026 – Nov 17, 2026

#2 in Claude Codeself-reported
sessions
2,288
tokens
4.7B
active days
65
longest streak
45 days
estimated spend
$4,383

leaderboard ranks

earned titles

  • The Clear Briefer

    0.00 mid-run corrections per session (0.10 or fewer earns it)

  • The Architect

    58% of sessions that edited files started with a plan (half or more earns it)

  • The Context Keeper

    0% of sessions hit a context reset (5% or fewer earns it)

agents, by share of sessions

  • Claude Code 99%
  • Codex 1%

fun factRan 8 agents at once.

how they work

ratios of counts, not volume
Mid-run corrections
0.00
Sessions with a context reset
<1%
Tool-call error rate
3%
Started with a plan
58%
Sessions at once
1.2×

Recap cards

16 cards
  • Persona

    The Golden-Hour Orchestrator

    2,288 sessions in 82 days

  • Most agents at once

    8 at once

    sessions side by side, plus 125 subagents

  • Tokens processed

    4.7B

    tokens, about $4,383 at API list prices

  • Longest streak

    45 days

    in a row, Oct 4 to Nov 17

  • Longest session

    7 hrs

    of active work in one Claude Code session

  • When they prompt

    Golden-hour builder

    busiest hour 17:00

  • Manners check

    6 pleases

    4 thank-yous and 5 sorries in 3,603 prompts

  • Danger zone

    85 × rm -rf

    0 force pushes and 47 naps. Nothing caught fire.

  • Time with agents

    130 hrs

    with an agent at work, 154 agent-hours in all

  • Top model

    Claude Opus 5.5

    answered 56% of 26,305 model requests

  • Favourite day

    Friday

    the busiest day, 609 prompts

  • Test runs

    7 runs

    of the tests, 100% passed

  • Bash hall of fame

    20,111

    shell commands run by agents

  • Where things broke

    811

    failed tool calls, 3% of 27,430

  • Words typed

    289.7K

    words in prompts, 15% of them questions

  • Slash commands

    4

    slash commands typed

Published by @rtuckute from their own AI Insights report. Self-reported, not verified. · see the leaderboard

Find out yours.

One command in your terminal. It reads your history on your machine and makes the recap there. Nothing is uploaded.

$ npx ai-insights@latest

Your report stays on your machine. Publishing a profile like this one is optional and sends only the lines you see.