io.github.Ezed9/mcp-bouncerai

Bouncer

Verified · today

Stdio MCP proxy that denies agent sends to destinations first seen in untrusted tool output.

Install

claude mcp add mcp-bouncer -- uvx bouncer-mcp

Our take

Proxies stdio MCP traffic and blocks agents from sending data to destinations first introduced through untrusted tool output. It addresses a specific exfiltration path for security-conscious MCP users, but its narrow rule should be treated as one defense layer rather than comprehensive destination or content security.

reviewed by hand · 2026-11-13

Something wrong or dead here? Report it

Verification record

last verified
today
github stars
6
last commit
today
archived
no
license
MIT
in registry since
2026-11-12

github.com/Ezed9/mcp-bouncer

More in ai

browse all →
reaai

REA

Verified · 27 days ago

Reverse engineer anything from your terminal or agent with one CLI and MCP server.

hand-reviewed291k stars232k dl/wkchecked 27 days ago

rufloai

Ruflo

Verified · 3 days ago

Agent teams, swarms, memory and hooks for Claude Code and Codex, served over MCP.

hand-reviewed81k stars72k dl/wkchecked 3 days ago

claude-flowai

claude-flow

Verified · 2 mo ago

AI orchestration with hive-mind swarms, neural networks, and 87 MCP tools for enterprise dev.

hand-reviewed73k stars19k dl/wkchecked 2 mo ago

tldrawai

tldraw

Verified · 2 mo ago

Draw and visually collaborate with your agents on tldraw's canvas.

hand-reviewed50k starschecked 2 mo ago

codebase-memory-mcpai

Codebase Memory

Verified · 2 mo ago

Codebase knowledge graph for AI agents — 162 languages, sub-ms queries, 99% fewer tokens.

hand-reviewed45k stars12k dl/wkchecked 2 mo ago