I built this after watching my Claude Code setup burn ~60,000 tokens just loading tool definitions from four MCP servers—before I even typed a prompt.
The problem: MCP gives agents access to hundreds of tools, but every tool description eats context. Redis published data showing the same thing—167 tools, 42% selection accuracy, 60K tokens of overhead per session. In production, it can hit 150K+ tokens. The agent spends more time deciding what to use than actually solving your problem.
Existing solutions fell into two buckets:
- Manual whitelists (mcpwrapped, MCP Funnel): You have to know in advance which tools to hide. With 100+ tools across multiple servers, that's a part-time job.
- Commercial platforms (Stacklok, Redis): Excellent accuracy (Stacklok hits 94%), but closed-source or require Redis infrastructure.
I wanted something that:
1. Works out of the box with my existing claude_desktop_config.json
2. Figures out which tools I need based on what I'm actually trying to do
3. Runs 100% locally, no API keys, no telemetry
4. Is open-source and dead simple to contribute to
So I built `shutup-mcp` in about 200 lines of Python. It aggregates tools from all your MCP servers, builds a local embedding index (all-MiniLM-L6-v2, ~80MB), and filters tools by intent before the agent sees them. Token reduction in my testing is ~98%—similar to what Redis and Atlassian reported with comparable approaches.
It's v0.1.0, so there's plenty left to build (dynamic re-filtering per message is next). I'd love feedback from anyone else hitting MCP tool sprawl—what's your current workaround? Regex filters? Specialized agents? Just eating the token cost?
Thanks for checking it out. Happy to answer any questions in the comments.
"Not a dashboard, a co-worker" this positioning resonates. As a solo founder, I don't have a data team to investigate why a metric dropped. I just see the drop and panic. The idea of an AI that can connect the dots between a code release and a metric change is exactly what small teams need. How granular does the root cause analysis get?
98% token reduction is wild. I've been working with MCP servers and the tool list bloat is a real problem the agent spends half its context just reading tool descriptions. The zero-config approach is smart because the moment you ask devs to configure which tools to expose, you've already lost. How does it decide which tools to surface for a given query?
About shutup-mcp on Product Hunt
“Zero-config MCP proxy that hides 99% of tools”
shutup-mcp launched on Product Hunt on April 14th, 2026 and earned 66 upvotes and 4 comments, placing #39 on the daily leaderboard. Zero-config MCP proxy that shows agents only the tools they actually need — 98% token reduction, no API keys. - hjs-spec/shutup-mcp
shutup-mcp was featured in User Experience (364.7k followers), Developer Tools (511k followers), Artificial Intelligence (466.2k followers) and GitHub (41.2k followers) on Product Hunt. Together, these topics include over 201.3k products, making this a competitive space to launch in.
Who hunted shutup-mcp?
shutup-mcp was hunted by yuqiang@JEP. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
Want to see how shutup-mcp stacked up against nearby launches in real time? Check out the live launch dashboard for upvote speed charts, proximity comparisons, and more analytics.