io.github.vaquarkhan/mcp-bastionai

MCP-Bastion

Verified · 35 days ago

Security middleware for MCP. Blocks prompt injection, PII leakage, and resource exhaustion.

Install

claude mcp add mcp-bastion -- uvx mcp-bastion-python

Our take

MCP-Bastion is security middleware for MCP, focused on blocking prompt injection, PII leakage, and resource exhaustion. It is for teams running agent systems who want defensive controls in front of tool execution. The problem area is important, but security claims in this space are hard to validate from metadata alone, and the current project signals suggest an early implementation rather than a battle-tested safeguard.

reviewed by hand · 2026-07-16

Something wrong or dead here? Report it

Verification record

last verified
35 days ago
github stars
5
last commit
35 days ago
archived
no
license
NOASSERTION
in registry since
2026-07-24

github.com/vaquarkhan/MCP-Bastion

More in ai

browse all →
claude-flowai

claude-flow

Verified · 17 days ago

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

hand-reviewed73k stars19k dl/wkchecked 17 days ago

tldrawai

tldraw

Verified · 18 days ago

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

hand-reviewed50k starschecked 18 days ago

codebase-memory-mcpai

Codebase Memory

Verified · 8 days ago

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

hand-reviewed45k stars12k dl/wkchecked 8 days ago

openmetadata-mcpai

OpenMetadata

Verified · today

Official OpenMetadata MCP: 21 read and write tools for search, lineage, data quality, governance.

hand-reviewed15k starschecked today

mobilerunai

mobilerun

Verified · 20 days ago

Control real Android and iOS devices with LLM agents — tap, swipe, type, automate flows.

hand-reviewed9.4k starschecked 20 days ago

praisonaiai

PraisonAI

Verified · 7 days ago

AI Agents Framework with Self Reflection and MCP support

hand-reviewed9.1k starschecked 7 days ago