Auto-retry Claude Code on subscription rate limits, API overload (529/5xx) and safeguard false positives — waits for the printed reset, exponential backoff, tmux-based
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add --transport http cheapestinference https://mcp--cheapestinference.run.tools~/.codex/config.toml
[mcp_servers.cheapestinference]
url = "https://mcp--cheapestinference.run.tools"
# codex mcp add cheapestinference --url https://mcp--cheapestinference.run.tools~/.cursor/mcp.json
{
"mcpServers": {
"cheapestinference": {
"url": "https://mcp--cheapestinference.run.tools"
}
}
}.vscode/mcp.json
{
"servers": {
"cheapestinference": {
"type": "http",
"url": "https://mcp--cheapestinference.run.tools"
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"cheapestinference": {
"httpUrl": "https://mcp--cheapestinference.run.tools"
}
}
}claude_desktop_config.json · Customize › Connectors › Add custom connector
{
"mcpServers": {
"cheapestinference": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp--cheapestinference.run.tools"
]
}
}
}.trae/mcp.json
{
"mcpServers": {
"cheapestinference": {
"url": "https://mcp--cheapestinference.run.tools"
}
}
}