ResearchMeasured on a real machine.
One developer's sessions, read by Caprock, with the method and the date beside every figure.
Research · updated
49% of my Claude Code tool calls ran commands
In 58 days of my Claude Code sessions, 49% of tool calls ran shell commands. The call counts, cost at API list prices and cache figures from one machine.
Research · updated
Claude Code analytics: what Anthropic shows a team
Anthropic gives a team a dashboard, a spend report, two APIs and OpenTelemetry. What each one shows, what it does well, and what none of them answer.
Research · updated
Claude Code cost per month: $5,898 on a $200 plan
My August of Claude Code on a $200 Max plan came to $5,898 at API list prices — 29 times the fee. Three months of usage from one machine.
Research · updated
Claude Code had already deleted half my history
It keeps transcripts for 30 days. On my machine that was 67 sessions and $7,808 of work — gone from disk, still searchable in Caprock.
Research · updated
Five coding agents, five formats, one table
Claude Code, Codex, OpenCode, Gemini CLI and DeepSeek Harness each record their work differently, and three of the differences give a believable wrong number.
See every coding agent you run.
Who is waiting on you, who is stuck, how close the limit is. Free and open source; nothing leaves your machine.
Notarized by Apple. One file for Apple silicon and Intel, macOS 13 or later.
Or run it yourself:
Paste in Terminal: it installs Caprock and opens it.Or with Homebrew: brew install --cask dspv/tap/caprock-app