What your agent setup costs before it runs

Every skill description, agent listing and memory index is re-sent at the top of every session and every scheduled run, whether or not anything uses it. This counts that floor and prices it at today's published rates.

Runs entirely in your browser · nothing is uploaded · open your network tab and check

Point it at your configuration

Choose the .claude folder in your HOME directory, which is where skills and scheduled tasks live: C:\Users\you\.claude on Windows, ~/.claude on macOS and Linux.

Most projects have a .claude folder too, and it is not the one you want: it holds settings and launch config, no skills and no routines. The folder is hidden on every platform, so if the dialog will not show it, type the path straight into the dialog's folder box.

It reads only definition files: skills, scheduled tasks, agents, commands, CLAUDE.md and the memory index. Transcripts, caches and node_modules are skipped and never opened.

or drop one here

Cannot pick a folder? Paste one file instead

Paste the contents of a single SKILL.md, agent or CLAUDE.md. Give it a path so it can be classified, for example skills/my-skill/SKILL.md.

This is the floor, not your bill. It is what your configuration costs before anything happens, and it is usually a tiny fraction of what you actually spend.

On one real install we measured a week of ordinary use at 959 million tokens. The configuration floor on that same machine was 0.03% of it. Almost everything was the conversation itself, re-sent on every turn.

So treat the figures below as the part you change by editing files. The rest you change by how you work: shorter sessions, less context carried, cheaper models on the steps that do not need a big one.

Loaded every session

tokens

Per month

Sessions a month

Priced on

What is in the floor

What to do about it

What the numbers show

What this cannot see

Stated up front rather than in a footnote, because a cost tool that implies it has counted everything is worse than no cost tool.

Tool schemas from MCP servers. These are usually the largest single item in a session's fixed cost, often several times everything measured here. They are produced by the servers when they connect rather than stored on disk, so nothing on your filesystem can be counted for them.

Context growth inside a session. This measures what a session starts with, never what it accumulates as work goes on, and the second is normally much larger. For what actually happened rather than what a configuration costs, price a real usage export in the bill auditor.

Anything a plugin brings with it. A plugin marketplace on disk is a catalogue of skills and agents you could install, and nothing there says which are actually installed and loading. Counting the catalogue overstated one real setup by 31 skills and 34 agents, so it is excluded entirely. If you run installed plugins, your real floor is larger than the figure above, not smaller.

Whether any of it is worth keeping. This reports what text costs and where the same text is paid for twice. It has no opinion on whether a description is too long or a routine too frequent, because that is a question about your work and we publish no quality judgments about anything.

Tokens are estimated at four characters each, the same estimator used across this site. Real tokenisers differ by a few percent per provider, so treat these as close rather than exact. Rates come from our own price archive.