You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When you connect MCP servers to Claude Code, every tool definition gets injected into the system prompt on every API call. If you have 5 MCP servers with 40 tools, those definitions consume input tokens on every single turn -- whether you use the tools or not.
This cost is completely invisible today. Nobody tracks it.
What CodeBurn could show
Total token cost spent on MCP tool definitions per day/week/month
Which MCP servers contribute the most overhead
Which tools you actually used vs. which ones just sat in the system prompt burning tokens
A ratio like: "You paid $12.40 this week for MCP tool definitions, but only used 3 of 40 tools"
Why this matters
For heavy MCP users, tool definition overhead could be a significant chunk of their bill with zero visibility. This would be a unique feature -- no other tool surfaces this data.
How it would work
Claude Code session data already contains the full system prompt and tool definitions per turn. CodeBurn can parse these, measure token counts per tool definition, and cross-reference against actual tool_use blocks to calculate used vs. unused overhead.
Vote
Upvote this discussion if you want this feature. We will build it if there is enough demand.
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
Uh oh!
There was an error while loading. Please reload this page.
-
The problem
When you connect MCP servers to Claude Code, every tool definition gets injected into the system prompt on every API call. If you have 5 MCP servers with 40 tools, those definitions consume input tokens on every single turn -- whether you use the tools or not.
This cost is completely invisible today. Nobody tracks it.
What CodeBurn could show
Why this matters
For heavy MCP users, tool definition overhead could be a significant chunk of their bill with zero visibility. This would be a unique feature -- no other tool surfaces this data.
How it would work
Claude Code session data already contains the full system prompt and tool definitions per turn. CodeBurn can parse these, measure token counts per tool definition, and cross-reference against actual
tool_useblocks to calculate used vs. unused overhead.Vote
Upvote this discussion if you want this feature. We will build it if there is enough demand.
Related: #2
Beta Was this translation helpful? Give feedback.
All reactions