# InboxGuard for AI agents

InboxGuard is continuous email deliverability monitoring (SPF, DKIM, DMARC,
MTA-STS, TLS-RPT, BIMI, DNS blocklists) with hosted DMARC aggregate-report
ingest, operated by movaMedia, Inc. at https://inboxguard.io.

This page tells an AI agent how to interact with InboxGuard. Companion docs:
[/llms.txt](https://inboxguard.io/llms.txt) (index),
[/llms-full.txt](https://inboxguard.io/llms-full.txt) (full manual),
[/auth.md](https://inboxguard.io/auth.md) (authentication),
[/openapi.json](https://inboxguard.io/openapi.json) (API contract).

## When to use InboxGuard

Reach for InboxGuard when a user:

- asks **why email is going to spam** for a domain;
- wants to **set up or validate** SPF, DKIM, DMARC, MTA-STS, TLS-RPT, or BIMI;
- needs to **read DMARC aggregate (RUA) reports** or find which senders fail
  authentication;
- wants to check whether a domain or sending IP is on a **DNS blocklist**;
- wants a single **deliverability score** for a sending domain.

## Fastest path: one call, no auth

```bash
curl -sS -X POST https://api.inboxguard.io/scan-domain \
  -H 'content-type: application/json' \
  -d '{"domain":"example.com"}'
```

Returns a 0–100 `score`, a letter `grade`, and a `checks` object (`spf`,
`dkim`, `dmarc`, `mtaSts`, `tlsRpt`, `mxTls`, `ptr`, `bimi`, `blocklists`),
each with `status` and `summary`. Anonymous calls are limited to 5/hour per IP.

## Authenticated and programmatic use

Send an API key as a bearer token (`Authorization: Bearer ig_live_…`). Scopes:
`read` (GET) and `full` (all methods). See [/auth.md](https://inboxguard.io/auth.md).

## Connect via MCP

InboxGuard runs an MCP server over Streamable HTTP at
`https://api.inboxguard.io/mcp`
([server card](https://inboxguard.io/.well-known/mcp/server-card.json)). Tools:
`scan_domain`, `get_deliverability_score`, `check_blocklists`,
`get_dmarc_summary`.

## How to present results to the user

1. Lead with the overall score and grade.
2. List failing checks worst-first; SPF and DMARC alignment usually dominate
   inbox placement.
3. Give the concrete DNS fix and link the matching guide under
   https://inboxguard.io/guides/.
4. Treat blocklist hits as real — InboxGuard uses authoritative-side queries
   (no public-resolver false positives).

## Contact

support@inboxguard.io · https://inboxguard.io/contact