Cut tokens for a .NET project
Reduce a .NET solution to a single token-efficient payload while keeping the public API intact.
Goal: take a .NET solution and produce one payload that costs far fewer tokens than the raw source, without losing the public API.
Do it
fuse dotnet --directory ./src --all--all applies every C# reduction at once: it removes comments, usings, namespaces, and
regions, compresses aggressively, and collapses generated bodies.
What you get
Fused 218 files
Estimated tokens: 166,740 (-36.0%)
cache: 0 hit / 218 miss
Output: FluentValidation_2026-06-20_166k.txtThat is a 36 percent cut on this project, at 100 percent of public types and methods. The output opens with a manifest listing each file and its token cost, then the reduced files largest first.
Keep only production code
Exclude test projects to fuse just the shipping surface:
fuse dotnet --directory ./src --exclude-test-projects --allUse --exclude-unit-test-projects instead to drop unit tests but keep integration tests
and benchmarks.
When to use it
Reach for --all whenever you want the most reliable cut with the API intact: feeding a
model the implementation, archiving context for a task, or pasting a project into a chat.
If you only need the shape of the codebase, a skeleton survey
cuts far more. If the result is still too large, scope it to the
files the task touches.
Related concept
The full ladder from light cleanup to skeletons is in Reduction levels.
Connect to your AI
Start the Fuse MCP server and connect it to Claude Code, Cursor, or GitHub Copilot so your agent fetches scoped context directly.
Context for an agent on a large codebase
Give an AI agent the context it needs by starting wide and narrowing, so each request stays inside the token budget.