Catch the bad plan before you spend the tokens. Cross-LLM peer review via MCP.
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add inspectrum -- npx -y inspectrum~/.codex/config.toml
[mcp_servers.inspectrum]
command = "npx"
args = ["-y", "inspectrum"]
# codex mcp add inspectrum -- npx -y inspectrum~/.cursor/mcp.json
{
"mcpServers": {
"inspectrum": {
"command": "npx",
"args": [
"-y",
"inspectrum"
]
}
}
}.vscode/mcp.json
{
"servers": {
"inspectrum": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"inspectrum"
]
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"inspectrum": {
"command": "npx",
"args": [
"-y",
"inspectrum"
]
}
}
}claude_desktop_config.json
{
"mcpServers": {
"inspectrum": {
"command": "npx",
"args": [
"-y",
"inspectrum"
]
}
}
}.trae/mcp.json
{
"mcpServers": {
"inspectrum": {
"command": "npx",
"args": [
"-y",
"inspectrum"
]
}
}
}