transactional-email-setup
Skill0 starsChanged 50 days agoMITView on GitHub
Use when the user wants to send application email (password resets, receipts, notifications, sign-up confirmations) through Warmerly's transactional email API, verify a sending domain for it, check why a transactional email did not arrive, or send one through the Warmerly MCP server.
In this repository3 other files it gives its agents.
- deliverability-doctorSkill
- launch-outreach-campaignSkill
- reply-triageSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 72
If send_transactional_email is not available on this connection, have the user send a test from the dashboard at …
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Transactional email setup
Warmerly's transactional email is for mail a person expects because they did something: a password reset, a receipt, an alert. It runs on separate mail servers from cold outreach and warmup so that one can never damage the other. Keep it that way.
Guardrails (read first)
- Never send cold or marketing email from a transactional domain. Not once, not "just a
small list". Transactional mail rides on its own IP reputation; a single cold blast can push
every password reset the user's product sends into spam. Cold outreach belongs in a campaign,
from a separate domain (see the
launch-outreach-campaignskill). - You send real email.
send_transactional_emaildelivers to a real inbox. Unless the user has explicitly given you autonomy, confirm the recipient and content before sending, and never loop it over a list of addresses. - Respect plan limits and the suppression list. Suppressed recipients are dropped automatically; do not try to work around that by changing the address.
Step 1: What exists
list_sending_domains: each domain with its status. active means verified and ready to send
from. pending_dns means records are not published yet. No domains means start at Step 2. If the
user's workspace does not have transactional email enabled yet, the dashboard shows "early access".
Step 2: Add and verify a domain
Adding a domain happens in the dashboard at https://app.warmerly.com/transactional (Domains tab).
Recommend a subdomain dedicated to this job, such as notify.example.com or mail.example.com,
rather than the root domain, and never a domain the user already uses for cold outreach or for
Warmerly OneMail mailboxes (Warmerly refuses a domain that already has another job).
Warmerly shows three records to publish at the user's DNS provider:
- SPF: an
include:for Warmerly's transactional servers. If the domain already has an SPF record, add the include to that one record; do not publish a second SPF record. - DKIM: a TXT record with the key Warmerly generated. Copy it exactly, as one string.
- DMARC: a
_dmarcTXT record. Any valid DMARC record passes verification; recommend including arua=report address so failures are visible.
Then press Verify in the dashboard. All three present makes the domain active. Use
check_domain on the domain to confirm what the world sees if verification keeps failing. DNS
changes can take from minutes to a few hours to appear.
MX advice. Warmerly does not add an MX record for a transactional domain: the user keeps
receiving mail wherever they do today. But a domain that cannot receive mail at all looks like a
throwaway spam domain to inbox providers. If check_domain shows no MX for the sending domain or
its parent, recommend adding one (the user's normal mail provider's MX, so replies reach a real inbox).
Step 3: Send
In the user's application, the send goes through the transactional REST API with a transactional
API key (created in the dashboard's API keys tab), and every send should carry an
Idempotency-Key header unique to that event (for example password-reset:<userId>:<token id>).
A retry with the same key never sends twice: it returns the first message and marks it a duplicate.
search_docs for "transactional" finds the full API reference.
From here, send_transactional_email sends one message directly, which is useful for a test or a
one-off. The From address must be on an active domain. A good test email:
- has a real subject like "Your sending domain notify.example.com is connected", not "test";
- has both a text and an HTML part and a line saying why it was sent;
- goes to an address the user controls.
The tool makes a retry of an identical send safe on its own. If it reports that a send may have been interrupted, do not send again blindly: look it up first (Step 4).
If send_transactional_email is not available on this connection, have the user send a test from
the dashboard at https://app.warmerly.com/transactional.
Step 4: Did it arrive?
list_transactional_emails (filter by status, for example bounced or failed) and
get_transactional_email for the event timeline of one message:
sent: accepted by Warmerly's server. Not yet proof of delivery.delivered: the recipient's server accepted it. If the user still cannot find it, it is in spam or a filter: check the domain withcheck_domainand the content (one-line "test" emails get filtered).deferred: the recipient's server said try later; Warmerly retries.bounced: permanent failure; the reason is in the timeline. The address is suppressed afterwards.suppressed: the recipient was on the suppression list, so nothing was sent.failed: the submission failed. The timeline says whether it may have gone out.
When a tool is missing
send_transactional_email may not be available yet. Domain setup and API keys are always done in
the dashboard at https://app.warmerly.com/transactional; nothing in this workflow needs a password
or key to pass through you, and you should never ask the user to paste an API key into the chat.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

