Self-hosted LLM gateway and AI context firewall for PII/PHI containment, prompt-injection defense, and governed agent tools.
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add yagami-ai-context-firewall -- uvx yagami~/.codex/config.toml
[mcp_servers.yagami-ai-context-firewall]
command = "uvx"
args = ["yagami"]
# codex mcp add yagami-ai-context-firewall -- uvx yagami~/.cursor/mcp.json
{
"mcpServers": {
"yagami-ai-context-firewall": {
"command": "uvx",
"args": [
"yagami"
]
}
}
}.vscode/mcp.json
{
"servers": {
"yagami-ai-context-firewall": {
"type": "stdio",
"command": "uvx",
"args": [
"yagami"
]
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"yagami-ai-context-firewall": {
"command": "uvx",
"args": [
"yagami"
]
}
}
}claude_desktop_config.json
{
"mcpServers": {
"yagami-ai-context-firewall": {
"command": "uvx",
"args": [
"yagami"
]
}
}
}.trae/mcp.json
{
"mcpServers": {
"yagami-ai-context-firewall": {
"command": "uvx",
"args": [
"yagami"
]
}
}
}