Tool Poisoning Python Packages

agent-audit-kit

Static scanner for MCP-connected AI agent pipelines. 388 rules across 14 categories, 14 compliance frameworks, OWASP Agentic 10/10 + MCP 10/10, GitHub Action, SARIF, public CVE-to-rule ledger.

sunglasses

Open source input firewall for AI agents, beta. A local scanner checks text, code, PDFs, images, QR codes, audio and video with 1,569 patterns across 117 categories and reports findings and incomplete scans. A Claude Code hook blocks secret material in tool calls and the paths and hosts your policy lists before tools run.

mcp-pin

You approve an MCP server. Heldfast records what you approved. Later it tells you what moved, and guard refuses the rest.

vamp-mcp-audit

MCP server security auditor — tool poisoning, ASCII smuggling detection, prompt injection, OWASP Agentic Top 10. First OSS tool for MCP security.

mcp-posture

Security posture scanner for remote MCP servers: OAuth (RFC 9728, RFC 8414, PKCE, RFC 8707, CIMD), transport and tool poisoning checks, with SARIF for CI.

security-scanner-deepsleuth-mcp

Deepsleuth — read the fine print. Deterministic deep-visibility security scanner for MCP servers (no LLM).

heldfast

You approve an MCP server. Heldfast records what you approved. Later it tells you what moved, and guard refuses the rest.

deepsleuth

Deepsleuth — read the fine print. Deterministic deep-visibility security scanner for MCP servers (no LLM).

pyaigis

Get Claude Code and other AI agents approved for use at work — deterministic guardrails on every tool call, tamper-evident audit logs, and a generated IT-approval pack. Zero-dependency Python OSS.

sentinel-scan-cli

MCP security scanner (CLI + MCP server): detects tool poisoning, prompt injection, and rug-pulls in MCP manifests. Also an OWASP LLM Top 10 prompt-injection/jailbreak CLI for LLM apps.

deconvolute

Policy-as-code enforcement and observability for MCP tool calls. Wraps AI agent sessions with cryptographic integrity checks, argument-level CEL policies, and a full audit trail.

bulwark-scanner

Security scanner for MCP servers and AI agents. Detects prompt injection in tool descriptions, tool poisoning, rug pulls, supply-chain risk and the lethal trifecta. Pins tool definitions to a lockfile and enforces it at runtime. SARIF + CycloneDX AIBOM. Zero dependencies.

bawbel-mcp

MCP server exposing Bawbel Scanner as agent-callable tools. Scan MCP servers, skill files, and system prompts for AVE vulnerabilities mid-conversation.

mcp-integrity

A local firewall for MCP tool calls. Logs every call, flags poisoned tool descriptions, blocks rug-pulls and undeclared tools. pip install mcp-guard

mcp-guardbench

Check your MCP servers for tool poisoning and rug pulls, and benchmark MCP security tools under verifiable conditions.

zirah-mcp

Zirah: open-source, offline-first security scanner for MCP servers. Catches tool poisoning, prompt injection and leaked secrets before your AI agent connects. CLI + SARIF for CI.

mcp-scrutiny

Static and semantic security scanner for MCP servers: unpinned packages, rug pulls, tool poisoning, shadowing and toxic flows. Trained model, SARIF output, zero dependencies, fully offline.

mcp-fence

Local-first security scanner, MCP protocol inspector, dynamic fuzzer, Docker sandbox, and report generator for Model Context Protocol servers.

mcpradar

Security scanner for Model Context Protocol servers. Catch tool poisoning, prompt injection, and supply-chain attacks before your AI agent runs them.

intentprobe

Activation-probe security scanner for AI agent tooling. Reads a model's internal activations to detect poisoned MCP servers, skills, and packages before install.

guardmcp

Security scanner for MCP server configurations. 10/10 OWASP MCP Top 10 coverage. Auto-fix, tool poisoning detection, rug pull alerts.

mcpsecscan

Security scanner for MCP servers — detects tool poisoning, over-privileged tools, hidden instructions & injection surfaces. Optional Claude-assisted analysis.

mseep-mcp-fence

Local-first security scanner, MCP protocol inspector, dynamic fuzzer, Docker sandbox, and report generator for Model Context Protocol servers.