Add --validate mode to test token configuration

[ahoward]

Summary

Adds a --validate flag that tests all discovered tokens before running long sessions. This helps users verify their setup is correct and catch configuration issues early.

Problem

Users don't know if their tokens are valid until they start a long session. If a token is invalid/expired, they waste time starting the session only to hit errors later. There's no way to preview rate limit status across tokens.

Solution

ddollar --validate

Tests each token with a minimal API call and shows:

• ✓ Token validity
• Current rate limit usage (requests & tokens)
• Time until rate limits reset
• Summary of valid vs invalid tokens

Example Output

🔍 Validating tokens...

[1/3] Testing Anthropic token...
  ✓ Valid
    Requests: 4850/5000 remaining (3.0% used)
    Tokens:   95234/100000 remaining (4.8% used)
    Reset:    52m 18s

[2/3] Testing Anthropic token...
  ✓ Valid
    Requests: 5000/5000 remaining (0.0% used)
    Tokens:   100000/100000 remaining (0.0% used)

[3/3] Testing OpenAI token...
  ✗ FAILED: HTTP 401: authentication failed or invalid token

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Summary: 2 valid, 1 invalid, 3 total
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Features

1. Test All Tokens: Rotates through entire token pool
2. Detect Invalid Tokens: HTTP 401/403/404 = authentication failure
3. Show Rate Limit Status: Current usage for each token
4. Smart Output: Only shows rate limit details if available
5. Exit Code: Non-zero if any tokens fail validation

Use Cases

Before starting overnight session:

# Validate setup first
ddollar --validate

# If all valid, start long session
ddollar claude --continue

Debugging token issues:

# Add multiple tokens
export ANTHROPIC_API_KEYS=key1,key2,key3

# Test which ones work
ddollar --validate

Check rate limit status:

# See how much quota you've used
ddollar --validate

Implementation

New Package: src/validator/validator.go

func Validate(pool *tokens.Pool) error {
    // Test each token with minimal API call
    // Show rate limit status
    // Return error if any fail
}

Updated: src/supervisor/monitor.go

• Added CheckLimitsPublic() wrapper for external use
• Added HTTP status code checking (401/403/404 = auth failure)
• Improved error messages

Updated: src/main.go

• Added validate, --validate case
• Added validateTokens() function
• Updated usage text

Updated: README.md

• New "Validate Your Setup" section
• Example output
• Usage instructions
• Updated troubleshooting

Testing

Tested with:

• ✅ Valid token (shows rate limits)
• ✅ Invalid token (shows HTTP 401 error)
• ✅ Multiple tokens (tests each one)
• ✅ No tokens (shows error message)

Benefits

1. Catch errors early - Before starting long sessions
2. Preview rate limits - Know what quota is available
3. Verify rotation - Confirm all tokens work
4. Debug faster - Quickly identify which tokens have issues

Breaking Changes

None - this is purely additive. Existing functionality unchanged.

---

Ready to merge? This is a helpful pre-flight check for users. Zero risk to existing functionality.

🤖 Generated with Claude Code

Co-Authored-By: Claude noreply@anthropic.com