frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Ask HN: How do you handle marketing as a solo technical founder?

15•lazarkap•2h ago•11 comments

Zooming UIs in 2026: Prezi, impress.js, and why I built something different

76•tinchox6•4h ago•34 comments

Tell HN: Anthropic no longer allowing Claude Code subscriptions to use OpenClaw

1090•firloop•3d ago•826 comments

Upwork Inc. violates its own DMARC and SPF policy

8•tmcdos•12h ago•7 comments

Ask HN: How are you orchestrating multi-agent AI workflows in production?

4•swrly•10h ago•3 comments

Ask HN: Any Interesting Niche Hobbies?

17•e-topy•1d ago•15 comments

Training mRNA Language Models Across 25 Species for $165

147•maziyar•5d ago•42 comments

Ask HN: Is there any interest in a native Qt/C++ Discord client?

36•txtsd•4d ago•36 comments

Ask HN: SoTA of Context Building Methods

4•h4ch1•12h ago•1 comments

Ask HN: Who wants to be hired? (April 2026)

90•whoishiring•5d ago•383 comments

Ask HN: Who is hiring? (April 2026)

277•whoishiring•5d ago•386 comments

Ask HN: Where are all the disruptive software that AI promised?

18•p-o•1d ago•17 comments

Claude Code Down

79•theahura•8h ago•72 comments

Ask HN: Learning resources for building AI agents?

10•7e10•2d ago•5 comments

Write Your Own Copy

30•operatingthetan•1d ago•16 comments

Ask HN: How do systems (or people) detect when a text is written by an LLM

41•elC0mpa•11h ago•60 comments

Microsoft Discontinuing Publisher. Alternatives?

16•supliminal•1d ago•15 comments

Ask HN: I don't get why Anthropic is limiting usage

6•ud0•1d ago•6 comments

Ask HN: LLM-Based Spam Filter

4•michidk•1d ago•1 comments

Intelligence Cannot Be Trained?

9•hyperzzw•2d ago•4 comments

Ask HN: How Do You Relax?

20•azeirah•2d ago•25 comments

Anthropic to limit Using third-party harnesses with Claude subscriptions

19•guiyuwei•3d ago•8 comments

Third-party Claude harnesses will now draw from extra usage

15•iBelieve•3d ago•5 comments

Skills to construct and read docs following the progressive exposure principle

5•Felix-Zhenghao•1d ago•1 comments

Claude Peptides – slash commands to cut Claude Code token usage by 73%

5•alchaka•1d ago•3 comments

Ask HN: What dev tools do you rely on that nobody talks about?

49•crcsmnky•5d ago•39 comments

Stripe closed my UAE business account and is withholding $3.5K

27•alganzory•4d ago•7 comments

Ask HN: How relevant is low code according to you in today’s world with Claude?

10•kinj28•4d ago•7 comments

Lightweight IDE to Pair with Claude Code?

8•zupancik•3d ago•2 comments

Ask HN: What is your dev set up like?

15•break_the_bank•4d ago•28 comments
Open in hackernews

Upwork Inc. violates its own DMARC and SPF policy

8•tmcdos•12h ago
I am not sure whether it happens on all outgoing emails or only on some of them. The SPF policy for upwork.com specifies that mail.clinchtalent.com and all IP addresses that are listed by spf.mandrillapp.com are allowed to send email on behalf of upwork.com

However, at least some (if not all) of the system emails that are generated and sent by the Upwork marketplace go through MailGun - and their IP addresses are missing from the SPF policy for upwork.com Additionally, the DMARC policy for upwork.com is set to "strict" - which means that if the SPF check fails then all RFC-compliant SMTP servers should reject the message.

I raised a support ticket and clearly explained the situation. The support agent admitted that he is not trained on such things and does not understand the overly technical part of my explanations (including screenshots and logs) - so I naturally asked for escalation to someone who is more qualified.

Quite expectedly, my request was ignored and we continued our conversation back and forth. I tried to explain the security and deliverability implications of such DNS misconfiguration for the Upwork company - and my words were again ignored.

Another support agent stepped-in (perhaps another shift) and we are back on step 1 - the situation is better than chatting with an AI but apparently not so much if unqualified staff refuses to transfer the ball to their more qualified colleagues.

I can understand that engineers do not want to be bothered with trivial things. But when the first line of support does not understand what I am talking about and we are exchanging a dozen of messages while a mid-level engineer would have got the thing already on step 1 - all the consequences go to the company first and then on its customers.

Comments

KomoD•11h ago
> The SPF policy for upwork.com specifies that mail.clinchtalent.com and all IP addresses that are listed by spf.mandrillapp.com are allowed to send email on behalf of upwork.com

No, it also lists Valimail as being able to make decisions on SPF. That's what the "include:%{i}._ip.%{h}._ehlo.%{d}._spf.vali.email" part is.

https://support.valimail.com/en/articles/8466461-valimail-sp...

tmcdos•11h ago
According to https://tools.sendmarc.com/spf-policy-test/upwork.com/198.24... v5142.v530814cf.use4.send.mailgun.net or c66.c5341538.usw1.send.mailgun.net are not allowed to send emails on behalf of upwork.com You can also check through https://spf.access.nu/ or https://dmarcian.com/spf-survey/ that IPs belonging to MailGun are not allowed to send emails for upwork.com
KomoD•10h ago
Those tools aren't using the macro which means they are not following the RFC, stop using crappy online tools and wasting people's time.

You can read about it here: https://datatracker.ietf.org/doc/html/rfc7208#section-7

dig +short TXT "159.112.254.142._ip.v5142.v530814cf.use4.send.mailgun.net._ehlo.upwork.com._spf.vali.email"

"v=spf1 include:mailgun.org -all"

--

dig +short TXT mailgun.org

"v=spf1 include:_spf.mailgun.org include:_spf.eu.mailgun.org -all"

--

dig +short TXT _spf.mailgun.org

"v=spf1 include:_spf1.mailgun.org include:_spf2.mailgun.org ~all"

--

dig +short TXT _spf2.mailgun.org

"v=spf1 ip4:104.130.122.0/23 ip4:146.20.112.0/26 ip4:161.38.192.0/20 ip4:143.55.224.0/21 ip4:143.55.232.0/22 ip4:159.112.240.0/20 ip4:198.244.48.0/20 ip4:204.220.168.0/21 ip4:204.220.176.0/20 ~all"

And there's 159.112.240.0/20.

--

The SPF lookup limit is 10 which means that this way of doing it is totally valid.

And here's where you can read about the lookup limit: https://datatracker.ietf.org/doc/html/rfc7208#section-4.6.4

tmcdos•9h ago
Got it. Thanks and apologies.
tmcdos•11h ago
After some investigation, it looks like only mailgun.org is declared in ValiMail but not mailgun.net, e.g. a DNS query for 198.244.56.66._ip.c66.c5341538.usw1.send.mailgun.net._ehlo.upwork.com._spf.vali.email returns "v=spf1 include:mailgun.org -all"
avian•3h ago
No idea about Upwork, but I had about the same situation about some other company sending me mail I cared about for a reason and their mail was not getting delivered to me because their DMARC check was failing.

They said "thanks, we'll look into it" and kept sending broken mail for years.

My guess is if you're a big enough player Google learns to ignore your broken DMARC config or somebody knows somebody on the inside who can add an exception. And then your mail gets delivered to @gmail.com just fine and that means it's working and wtf is this guy complaining about.

winstonwinston•2h ago
> Additionally, the DMARC policy for upwork.com is set to "strict" - which means that if the SPF check fails then all RFC-compliant SMTP servers should reject the message.

There is no “strict” policy. DMARC policy can be one of the following p= {none, quarantine, reject}.

The receiver decides if it wants to apply published DMARC policy for unauthenticated mail. What problem are you seeing exactly?

Remember both SPF and DKIM are used for policy evaluation.