투명한 MCP 프록시는 거의 모든 다른 MCP 서버 imaginable을 위한 큰 상류 공구 응답의 토큰 비용을 감소시킵니다. 단일 업스트림 MCP를 포장하고, 모든 통화를 전달하고, 콘텐츠 인식으로 지방 페이로드를 처리하는 것은 에이전트가 유니버셜 동사 ([peek], slice, aggregate], describe]], list_handles]])의 작은 세트로 쿼리 할 수 있습니다.

벤치 마크

두 개의 인기있는 Real-world MCP 서버에 대한 end-to-end를 측정했습니다. 각 실행은 도구의 고정 스크립트를 두 번 호출 — 한 번 직접, jmunch-mcp를 통해 한 번 - 세 개의 후속 jmunch.*] 부록 된 측에 동사 호출하여 에이전트가 전체를 슬러핑하는 것보다 큰 결과로 드릴링을 만듭니다.

| 스위트 | 업스트림 | 직접 토큰 | jmunch-mcp | 저장 | |—|—|—:|||||| | GitHub([T7]] 문제/PRs/commit) | [T8]] | 379,878 | 44,328 | 335,550 (88.3%) | | Firecrawl (Wikipedia scraps + 사이트 맵 + 검색) | firecrawl-mcp | 259,574 | 2,928 | 256,646 (98.9%) |

벽시 시간은 또한 추가 동사 통화에도 불구하고 두 스위트에 프록시로 빠졌습니다. 에이전트는 지방 페이로드를 통해 페이지가 없습니다.

| 스위트 | 직진 | jmunch-mcp | 델타 |

jmunch-mcp

Transparent MCP proxy that reduces the token cost of large upstream tool responses for nearly every other MCP server imaginable. Wraps a single upstream MCP, forwards every call, and handle-ifies fat payloads into content-aware backends the agent can query with a small set of universal verbs (peek, slice, search, aggregate, describe, list_handles).

Benchmarks

Measured end-to-end against two popular real-world MCP servers. Each run fires a fixed script of tool calls twice — once direct, once through jmunch-mcp — with three follow-up jmunch.* verb calls on the proxied side to model an agent drilling into a large result rather than slurping it whole.

Wall-clock time was also faster with the proxy on both suites, despite the extra verb calls — the agent never has to page through the fat payload:

Tabular content (GitHub) routes to the SQLite backend and answers peek/slice/aggregate; JSON content (Firecrawl scrape/map) routes to the JSON-tree backend and answers peek/slice (JSONPath)/search. See bench/README.md to reproduce.

Install

pip install jmunch-mcp

From source:

git clone https://github.com/jgravelle/jmunch-mcp
cd jmunch-mcp
pip install -e .

Quickstart

jmunch-mcp init

init scans three sources — your MCP client configs (Claude Desktop, Claude Code, Cursor, Windsurf, Continue), running processes, and a small catalog of popular upstreams (GitHub, Firecrawl, filesystem, fetch, Brave Search, Slack) — and renders a checklist. Tick the upstreams you want wrapped, and it writes one .toml per selection into ./configs/. Non-interactive flags: --yes (pick everything already registered in a client), --dry-run, --overwrite, --out , --no-running, --no-catalog.

Manual

jmunch-mcp --config examples/config.toml

Configure your MCP client to launch jmunch-mcp --config instead of the upstream server directly. Add --report to print a session summary on shutdown.

Gateway mode (v2 — universal proxy)

The MCP proxy above saves tokens for MCP clients. The gateway saves tokens for any AI application that speaks the OpenAI or Anthropic HTTP API — LangChain, LlamaIndex, CrewAI, AutoGen, Continue, Cline, Aider, or a raw SDK. No code changes in the app; just point base_url at jmunch.

pip install 'jmunch-mcp[gateway]'