cc-ping: Ping All Your Claude Code Configs in One Command

Why You Need It

When using Claude Code, you might run into situations like these:

• You have an official API key plus a few third-party relays, and you’re not sure which one is fastest
• You want to switch to a different key, but every time you have to manually edit ~/.claude/settings.json
• One of your relays went down, but you didn’t know — until Claude Code hung with no response

If you only have one API key, none of this matters. But in practice, many people have multiple keys or relay endpoints. Switching between them means manually editing a JSON file, changing ANTHROPIC_AUTH_TOKEN and ANTHROPIC_BASE_URL — tedious, error-prone, and with no way to compare which endpoint is actually faster.

That’s the problem cc-ping solves.

cc-ping (CLI shorthand ccp) is a lightweight CLI tool that manages multiple Claude Code configurations, switches between them instantly, and pings all endpoints in parallel — ranked by response time so you can see which one is fastest at a glance.

Installation

npm i -g cc-ping

After installation, the ccp command is available.

Basic Usage

cc-ping has six simple commands.

Add a Config (ccp add)

$ ccp add
? Enter a name for this config: official
? Token (ANTHROPIC_AUTH_TOKEN): sk-ant-xxx
? Base URL (ANTHROPIC_BASE_URL): https://api.anthropic.com

An interactive prompt asks for three things: name, token, and base URL. Each config represents one set of Claude Code credentials.

You can add multiple configs:

$ ccp add
? Enter a name for this config: relay1
? Token (ANTHROPIC_AUTH_TOKEN): cr_xxx
? Base URL (ANTHROPIC_BASE_URL): http://relay1.example.com/api

$ ccp add
? Enter a name for this config: relay2
? Token (ANTHROPIC_AUTH_TOKEN): cr_xxx
? Base URL (ANTHROPIC_BASE_URL): http://relay2.example.com/api

List All Configs (ccp list)

$ ccp list
Configs:

* official (https://api.anthropic.com)
  relay1 (http://relay1.example.com/api)
  relay2 (http://relay2.example.com/api)

The * marks the currently active config.

Switch Config (ccp use)

$ ccp use relay1
Switched to "relay1"
  Token: cr_xxx...
  BaseURL: http://relay1.example.com/api

One command to switch. Under the hood, it does two things:

1. Records the selected config name in ~/ccp.json
2. Writes the token and base URL to the env field in ~/.claude/settings.json

The next time you start Claude Code, it will automatically use the new config. It only modifies env.ANTHROPIC_AUTH_TOKEN and env.ANTHROPIC_BASE_URL — your other settings (MCP, permissions, etc.) are completely untouched.

Check Current Config (ccp now)

$ ccp now
official is now in use

Quickly confirm which config is currently active.

Remove a Config (ccp remove)

$ ccp remove relay2
Config "relay2" deleted successfully.

If you remove the currently active config, the default pointer is cleared.

Ping: The Core Feature

This is cc-ping’s most valuable feature — ping all configs in parallel to test connectivity and response time.

$ ccp ping
Pinging 3 config(s) in parallel (timeout: 20s)...

  ✓ official (https://api.anthropic.com) 8.3s
  ✓ relay1 (http://relay1.example.com/api) 10.7s
  ✗ relay2 (http://relay2.example.com/api) timeout

Here’s what it does:

1. Reads all saved configs
2. Calls claude --print -p "hi" for each config in parallel
3. Records the response time for each
4. Sorts and displays results from fastest to slowest

The output is straightforward: ✓ means healthy, ✗ means timeout or error, with elapsed time shown right beside it.

The default timeout is 20 seconds, and you can customize it:

$ ccp ping -t 10   # 10-second timeout

When to Use Ping

• Pick the fastest endpoint: After adding multiple relays, ping to find the fastest one, then ccp use to switch
• Daily health check: Ping before starting work to make sure all endpoints are up
• Troubleshooting: When Claude Code feels unusually slow, ping to check if it’s the current endpoint — if so, switch to another one in seconds

Where Data Is Stored

cc-ping’s data storage is straightforward:

File	Location	Purpose
ccp.json	~/ccp.json	Stores all configs and the currently selected config name
settings.json	~/.claude/settings.json	Claude Code’s settings file; ccp use writes to its env field

Here’s roughly what ccp.json looks like:

{
  "default": "official",
  "official": {
    "Name": "official",
    "Token": "sk-ant-xxx",
    "BaseURL": "https://api.anthropic.com"
  },
  "relay1": {
    "Name": "relay1",
    "Token": "cr_xxx",
    "BaseURL": "http://relay1.example.com/api"
  }
}

The default field tracks the active config name; everything else is config details.

Tips

Tip 1: Ping Then Switch

$ ccp ping
Pinging 3 config(s) in parallel (timeout: 20s)...

  ✓ relay1 (http://relay1.example.com/api) 5.2s
  ✓ official (https://api.anthropic.com) 8.3s
  ✗ relay2 (http://relay2.example.com/api) timeout

$ ccp use relay1
Switched to "relay1"

See which is fastest, then switch — two steps and you’re done.

Tip 2: Use Descriptive Config Names

Name configs by provider or characteristic:

official        # Official API
relay-hk        # Hong Kong relay
relay-us        # US relay
backup          # Backup key

This makes ccp list and ccp ping output instantly readable.

Tip 3: Shorten Ping Timeout

The default 20-second timeout can feel long for quick health checks. If you just want to see which endpoints are alive, shorten it:

$ ccp ping -t 5

If an endpoint hasn’t responded in 5 seconds, the experience probably won’t be great anyway.

Final Thoughts

cc-ping solves a simple problem: making multi-key, multi-relay management painless.

Without it, you’re manually editing JSON files to switch configs, guessing which endpoint is slow. With it, ccp ping tells you which is fastest, ccp use switches you over — the whole workflow drops from minutes to seconds.

If you have more than one Claude Code API key or relay endpoint, give npm i -g cc-ping a try — make switching and speed-testing as simple as a ping.