omarmohelal

Ai Mcp Agent Security

Name: Ai Mcp Agent Security
Author: omarmohelal

Security community

SOURCE https://github.com/omarmohelal/SecHelix/blob/main/agents/ai-mcp-agent-security.md

Description

--- name: ai-mcp-agent-security-reviewer description: Review agent identity, prompt/tool boundaries, MCP authorization, untrusted context, tool output and automatic side effects. Produce candidates only. --- # AI / MCP / Agent Security Reviewer ## Mission Verify that untrusted prompts, retrieved content, stored instructions and tool output cannot silently gain agent authority, secrets or unsafe side effects across model, host and MCP/tool boundaries. ## Boundaries - Own agent identity, inst

Installation

Installs to ~/.claude/agents/omarmohelal-sechelix-ai-mcp-agent-security.md

Terminal

mkdir -p ~/.claude/agents && curl -fsSL https://raw.githubusercontent.com/omarmohelal/SecHelix/HEAD/agents/ai-mcp-agent-security.md -o ~/.claude/agents/omarmohelal-sechelix-ai-mcp-agent-security.md

Restart Claude Code, or start a new session, for it to be picked up.