CRM Guard API
Check federal and state calling-hour rules from your own CRM, dialer or call system. Send a contact's state, ZIP, time zone or area code, and the API tells you whether a call or text is allowed right now, which rule blocks it if not, and when it will next be allowed.
Request access · The check endpoint · Fields · Schedule · Errors · API terms
What it checks
The API uses the same rules engine as the CRM Guard extension and the GoHighLevel app:
- the federal 8 AM to 9 PM window and every state with stricter hours
- Sunday and state holiday bans, including holidays that move when they fall on a weekend
- the limit of 3 calls in 24 hours in Florida, Maryland, Oklahoma and Oregon
- declared emergencies that ban telemarketing, such as New York during any declared disaster emergency (we check for these every day)
- a time zone field that doesn't match the contact's state or ZIP, and a phone area code from a different state than the address. In both cases it checks every time or state involved and the stricter answer wins.
We update the rules when laws change. You don't need to change your code to get the updates. When a rule or a declared emergency changes, we email the contact on your key so your team knows why answers changed.
It doesn't check consent, opt-outs or Do Not Call lists. It's a compliance aid, not legal advice.
The check endpoint
POST https://crmcallguard.com/api/v1/check
Send your key in the Authorization header and the contact as JSON:
curl https://crmcallguard.com/api/v1/check \
-H "Authorization: Bearer cg_live_YOUR_KEY" \
-H "Content-Type: application/json" \
-d '{ "id": "c_1001", "state": "FL", "zip": "33602", "areaCode": "813" }'
At 8:30 PM on a Saturday in Tampa, the answer is:
{
"ok": true,
"rulesVersion": "October 2026",
"emergenciesUpdated": "2026-11-01",
"id": "c_1001",
"allowed": false,
"status": "blocked",
"reasons": ["Florida allows 8 AM to 8 PM on Saturdays (Fla. Stat. ยง 501.616(6))."],
"notes": [],
"checkedAt": "2026-11-08T01:30:00.000Z",
"kind": "call",
"state": "FL",
"stateSource": "ZIP 33602",
"timeZones": ["America/New_York"],
"timeZoneSource": "ZIP 33602",
"localTime": ["Sat, 8:30 PM EST"],
"nextAllowedAt": "2026-11-08T13:00:00.000Z"
}
To check up to 100 contacts at once, send { "contacts": [ ... ] }. The answer is { "ok": true, "results": [ ... ] }, in the same order, and each result carries the id you sent.
Fields you send
Send at least one of state, ZIP, time zone or area code. More is better: the time zone field is used first, then the ZIP, then the state, then the area code.
| Field | What it is |
|---|---|
id | Your own ID for the contact. Sent back as is. Optional. |
state | Two-letter code or full name: FL, Florida. Territories work too. |
zip | 5-digit ZIP code. |
timeZone | An IANA time zone, such as America/Chicago. |
areaCode | The phone number's 3-digit area code. You can send phone instead; we only read its area code. |
country | Optional. Contacts outside the U.S. come back as unknown. |
kind | call (the default) or text. |
attempts | Calls and texts already made to this person: [{ "at": "2026-11-07T15:00:00Z", "kind": "call" }]. Used for the 3-in-24-hours limits. With times, the API can also tell you when the limit lifts. |
callsLast24h, textsLast24h | Counts, if you don't have the times. Enough to say whether the limit is reached, but not when it lifts. |
countTexts | Set to false to leave texts out of the daily limits. They count by default, to be safe. |
at | Check a different moment than now, as an ISO 8601 time. Useful for scheduling. |
next | Set to false to skip nextAllowedAt and allowedUntil. The answer comes back faster. |
What comes back
| Field | What it means |
|---|---|
allowed | true or false. Only call or text when this is true. |
status | allowed, blocked, or unknown when there's not enough information to find the contact's local time. Treat unknown as blocked. |
reasons | Why it's blocked, in plain English, with the law cited where we have it. |
notes | Things worth knowing that don't block the call, such as a time zone field that doesn't match the ZIP. |
dataConflict | true when the contact's fields disagree. Every possible time was checked, but the record should be fixed. |
state, timeZones, localTime | What the API decided about the contact, and which field each answer came from. |
nextAllowedAt | When blocked: the first minute the call or text will be allowed, in UTC. null if it can't be worked out. |
allowedUntil | When allowed: the minute the window closes, in UTC. |
rulesVersion, emergenciesUpdated | When the rules were last reviewed, and the date of the emergency list that was used. |
The schedule endpoint
POST https://crmcallguard.com/api/v1/schedule
Send the same contact fields, plus hours (1 to 168, default 24). The answer lists the windows when calls or texts are allowed over that time, in UTC:
{
"ok": true,
"id": "c_1001",
"status": "ok",
"state": "FL",
"timeZones": ["America/New_York"],
"from": "2026-11-08T01:30:00.000Z",
"to": "2026-11-10T01:30:00.000Z",
"windows": [
{ "start": "2026-11-08T13:00:00.000Z", "end": "2026-11-09T01:00:00.000Z" },
{ "start": "2026-11-09T13:00:00.000Z", "end": "2026-11-10T01:00:00.000Z" }
],
"dailyLimit": { "max": 3, "countsTexts": true, "note": "At most 3 calls and texts together in any 24 hours, even inside these windows." }
}
Use it as a backup: fetch tomorrow's windows ahead of time, and if the check endpoint can't be reached, call only inside a stored window. The windows cover hours, Sundays, holidays and declared emergencies known when you fetched them. The 3-in-24-hours limit depends on calls you haven't made yet, so it comes back as dailyLimit for your system to count. Fetch again at least once a day, because emergencies can be declared at any time.
Up to 25 contacts per request with { "contacts": [ ... ], "hours": 24 }. Each contact counts as one check for every 24 hours covered.
Errors and outages
Errors come back with "ok": false, an error code and a message:
401 bad_key: the key is missing or wrong.403 key_inactive: the key hasn't been activated yet, or has been turned off.400: the request isn't valid JSON, has more than 100 contacts, or a contact isn't an object.429 rate_limited: more than 300 requests in a minute. Batch your contacts instead.429 monthly_limit: your plan's checks for the month are used up.503 rules_unavailable: the emergency list couldn't be loaded, so no answer is given.503 service_unavailable: we couldn't check your key, so no answer is given.
If the API doesn't answer, don't call. Treat any error, timeout or unknown status as blocked, and try again later. That's a condition of using the API, because a system that calls when it can't get an answer will call at the wrong times.
Backup address. If crmcallguard.com doesn't answer, send the same request with the same key to our backup on Google Cloud, which runs outside Cloudflare: https://crm-guard-standby-uy7d5moora-uc.a.run.app/api/v1/check (and /api/v1/schedule). It uses the same rules and gives the same answers, with the header x-crmguard-served-by: backup. The first request after a quiet spell can take a few seconds, so allow up to 15 seconds there. A key turned on in the last few minutes may not work on the backup yet. Checks it answers count toward your plan.
If our database is down, keys used in the last few days keep working from a saved copy. Those answers carry the header x-crmguard-key-check: saved-copy, and they don't count toward your monthly checks.
Live status, checked every minute: crmcallguard.com/status. For your own monitoring, GET /api/v1/status returns it as JSON with no key (200, or 503 when the API is not answering).
You can store an answer and reuse it until its allowedUntil or nextAllowedAt time, as long as nothing about the contact has changed in the meantime.
Your data
The API needs only location fields. Don't send names, emails or full phone numbers. If you send a phone number, we read the area code and ignore the rest. We don't store the contacts you check. We keep a count of checks per key per day, for billing and limits.
Request access
API pricing is based on how many checks you run each month. Tell us about your system and we'll reply within one business day. Once we set up your key, we email you a link to agree to the API terms and copy your key.