Codex CLI
OpenAI Codex CLI passes the token through an environment variable (unlike Claude’s --header form): run export first, then codex mcp add:
export TWINKLE_HUB_API_KEY=sk-...
codex mcp add twinkle-hub --url https://api.twinkleai.tw/mcp/ \
--bearer-token-env-var TWINKLE_HUB_API_KEYIf Codex prints MCP client for twinkle-hub timed out after 30 seconds on first launch, the default 30-second startup timeout isn't enough for the cold start — open ~/ and add startup_timeout_sec = 60 under the [mcp_servers.twinkle-hub] block. Subsequent launches typically connect in under 5 s and you can lower it back:
[mcp_servers.twinkle-hub]
startup_timeout_sec = 60Pricing
Free allowance, rate limits and how the plans differ are all on the pricing page. When the allowance runs out, the gateway rejects calls until it refills, then resumes on its own — nothing to reset.
How to remove
Want to pause? Here's how to remove the Twinkle Hub MCP from each client. Your API key isn't revoked — you can re-add it any time.
Want to fully close your account?
The above only disconnects the client; your hub account and API key remain. To fully remove your account + data, go to /account/security and request deletion (72-hour grace period).