The latest articles on DEV Community by Cranius Maximus (@craniusmaximus). https://dev.to/craniusmaximus https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4064081%2Fa8fc173a-3050-4a6b-8d52-be16b0fcd3a0.png https://dev.to/craniusmaximus en Cranius Maximus Wed, 05 Aug 2026 14:09:29 +0000 https://dev.to/craniusmaximus/claude-code-already-shows-what-i-spent-i-wanted-what-never-fired-4cnn https://dev.to/craniusmaximus/claude-code-already-shows-what-i-spent-i-wanted-what-never-fired-4cnn <p>Claude Code already has <code>/usage</code> (same as <code>/cost</code>). On a Pro or Max plan it will show session cost estimates, plan bars, and a recent breakdown of usage by skill, subagent, plugin, and MCP server. </p> <p>But, I wanted to optimize Claude, or have it optimize itself, rather than tell me. Skills I installed months ago were sitting idle, and contributing tokens each turn. <code>/usage</code> is good at showing what <em>ran</em> and what counted against the plan. It does not answer: what is installed, still loads, and has not fired once across the last stretch of my own transcripts?</p> <p>That is what session-tax is for.</p> <h2> What it actually measures </h2> <p>It reads your Claude Code session transcripts off disk on your machine. No network calls in the free script. Imports are <code>fs</code>, <code>path</code>, and <code>os</code>.</p> <p>Across a window you choose (default 90 days) it reports:</p> <ul> <li>how many tokens a session typically opens with before you type (median and the worst tenth)</li> <li>installed skills versus skills that actually got invoked</li> <li>how many of those skills look dead (installed, not wired to an event, older than a fortnight, zero invocations in the window), and how many tokens their listings cost on every open</li> <li>counts for custom agents, connected servers, and hooks</li> </ul> <p>The dollar line is a rough guide. It prices opening context using the cache-read / cache-write mix it sees in your transcripts, and it says so. On a subscription you mostly pay this in overhead.</p> <h2> Free vs paid </h2> <p>Free edition: the totals above. It will tell you <em>how many</em> skills look dead and what that listing tax adds up to. It does not print the names.</p> <p>Paid ($19, one seat): names the dead skills, points at cold servers (connected but almost never called), machine-readable <code>--json</code>, and <code>--home</code> so you can audit a different profile. Shop link lives in the GitHub README on purpose. Some forums quietly filter <code>*.pages.dev</code> URLs.</p> <p>Repo: <a href="https://github.com/CraniusMaximus/claude-code-autonomy-pack" rel="noopener noreferrer">github.com/CraniusMaximus/claude-code-autonomy-pack</a></p> <h2> What I want to know </h2> <p>Does it help? Is it a waste of time? Any inaccuracies?</p> claude devtools productivity opensource