Agent Wrapped

agent-wrapped

**A wrapped card for how you actually work with Claude Code** — read from your own transcripts, rendered on your own machine, sent nowhere.

**[agent-wrapped.dev](https://agent-wrapped.dev)** · **[Changelog](https://github.com/lfgcampos/agent-wrapped/releases)**

npx @lfgcampos/agent-wrapped

Published as [`agent-wrapped`](https://www.npmjs.com/package/@lfgcampos/agent-wrapped). Earlier releases went out under `@lfgcampos/agent-wrapped`, which still resolves but is no longer updated — switch to the unscoped name to keep getting releases.

  CLAUDE CODE · 33 ACTIVE DAYS                  2026-07-16 → 2026-08-25

         394 : 1
         tokens read for every token written

  4-day streak · longest 18 · you work most at 3pm
  ▂▂▂▅▅▁▁▁▁▁▁▂▁▁▂█▆▇▃▄▅▆▆▃   19% of your writing is weekend work
  00                      23

   98%   of what it read was cache — the same context, re-sent
   42%   of your calls were subagents…
    6%   …but they wrote only that share of the words
   58%   of your writing went to 3 of your 19 repos

  WHAT YOU REACH FOR                    (% of 49,566 tool calls)
  ██████████████  62%  Bash
  ███░░░░░░░░░░░  14%  Read
  ██░░░░░░░░░░░░  13%  Edit
  █░░░░░░░░░░░░░   4%  Write

  15.2 agent turns for every message you send · 733 sessions

  BATTLE SCARS
     4   times you hit the usage limit, across 3 weeks
    14   times the server was overloaded on you

  agent-wrapped.dev

Add `--html` for a self-contained page at `~/.agent-wrapped/card.html`. It has a **Save as image** button that renders a 1200×630 PNG for sharing — drawn in your own browser, never uploaded anywhere.

Nothing leaves your machine

• No network calls. No server, no account, no analytics, no version check.
• Zero runtime dependencies. No third-party code to audit but this.
• It is about a thousand lines. Read src/ and confirm both claims yourself.

It reads exactly two paths, both read-only: `~/.claude/projects` (your transcript