llamaCVEVulnerability Intelligence
/
Advisory · GitHub · PyPI · eml-parser

GHSA-G6W4-PG2R-XX6W

Indicators written with HTML character references in e-mail body text are not reported

Texto original en inglés · Ver en español

Sin parche publicado
Descargar PNG JSON
CVSS 4.0
6.9
Media
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N Fuente: GitHub
ExplotaciónSin evidenciaNo figura en CISA KEV
Exploits públicos0Ninguno conocido
ParcheSin parcheEl fabricante no ha publicado corrección
Exposición3 dDesde la divulgación, 17 nov 2026, 11:33 a. m.
Detección1 dDesde la divulgación hasta su ingreso al índice
Timeline

Ciclo de vida de la vulnerabilidad

Hoy
Divulgación
Publicada +1
Explotación
Remediación
Monitoreo
Detectada +1
17 nov 26
18 nov 26
21 nov 26
17 nov 26
18 nov 26
19 nov 26
20 nov 26
21 nov 26
FechaFaseEventoDetalle
Divulgación Publicada GHSA-G6W4-PG2R-XX6W divulgada · CVSS 6.9 (GitHub)
Divulgación Modificada Última actualización del registro
Monitoreo Detectada Ingresó al índice
Monitoreo Verificada Última verificación contra las fuentes
Resumen
Descripción del advisory
Texto original · GitHub · en inglés

### Summary
eml_parser extracts indicators (URLs, domains, e-mail addresses, IP addresses) from the body of an e-mail. HTML character
references in the body text were not decoded before this search. An URL, domain, address or IP address written with
character references (for example `https://phishing.example.com/login`) is shown to the reader of an
HTML part as a normal link text, but eml_parser reported no indicator for it. The fix of CVE-2026-55618
(GHSA-fxgq-9m89-cxj9, version 3.0.2) only covered dots in an URL whose scheme was written as is.

### Details
Character references were decoded only in two places: inside matches of the URL pattern (which needs a literal
`scheme:`) and in HTML attribute values (`href`, `src`, ...). The indicator patterns themselves run on the raw text, and
the domain pattern deliberately rejects tokens touching an encoded dot. As a result, in text/html and text/plain parts:

  • `https://phishing.example.com/login` (and hexadecimal or named variants such as `.`,
  • `www.phishing.example.com`, `phishing.example.com`, `user@phishing.example.com` and
  • partly encoded values gave fragments instead: `user@phishing.example.com` was reported as the address

Version 4.7.1 decodes every text part once, the way its reader sees it, before the indicators are searched: HTML parts
by the rules of browsers (a legacy named reference without `;` in front of `=` or a letter or digit is kept, as in
attribute values), other text parts only for complete references ending in `;`, and attribute values by the rules for
attribute values. Nothing is decoded twice, so `:` stays `:`, as a browser shows it.

### Impact
Detection evasion. Systems that rely on the indicators reported by eml_parser (phishing triage, blocklist matching,
sandbox submission, incident response tooling) do not see a link or address that the recipient sees, so a malicious
message can pass checks based on these indicators. The parser does not crash and no other data is affected; the
recipient still has to follow the link for further harm, which is outside of eml_parser. Domains that are written
without references elsewhere in the message, for example in an `href` attribute or in the headers, were still reported.

### Workarounds
Decode HTML character references in the body text before passing it to an own indicator extraction, or apply an
additional extraction on the decoded text. There is no setting in eml_parser < 4.7.1 that changes this behaviour.

### Credits
Reported by @lanjianchun.

### References
- GHSA-fxgq-9m89-cxj9 / CVE-2026-55618 (earlier, incomplete fix in 3.0.2)
- Changelog of eml_parser 4.7.1

Vector CVSS 4.0

Desglose de métricas

Opción resaltada = valor del vector · izquierda = peor caso · GitHub
Vector de ataqueAV:N
RedAdyacenteLocalFísica
ComplejidadAC:L
BajaAlta
Requisitos de ataqueAT:N
NingunoPresentes
PrivilegiosPR:N
NingunoBajosAltos
InteracciónUI:N
NingunaPasivaActiva
Confid. (sistema)VC:N
AltoBajoNinguno
Integridad (sistema)VI:L
AltoBajoNinguno
Disponib. (sistema)VA:N
AltoBajoNinguno
Confid. (posterior)SC:N
AltoBajoNinguno
Integridad (posterior)SI:N
AltoBajoNinguno
Disponib. (posterior)SA:N
AltoBajoNinguno
Score de explotaciónQué tan fácil y probable es que la exploten
10.0/ 10Muy alto
Vector 10.0 (muy fácil) · sin evidencia de explotación.
Versiones

Rangos afectados

Afectada Corregida Última
eml-parser< 4.7.1
4.7.0
4.7.1
Remediación

Estado del parche por paquete

Paquete / productoEstadoCorregida enDesde divulgaciónÚltima versión
pypieml-parser
Sin parche — — 4.7.0 afectada
Versiones corregidas4.7.1
Explotación

Arsenal público

Sin exploits, módulos ni PoC públicos conocidos a la fecha de verificación.

Referencias
Advisoryhttps://github.com/GOVCERT-LU/eml_parser/security/advisories/GHSA-g6w4-pg2r-xx6w

Las referencias externas se muestran para copiar: cve.llama.ec no lo redirige a otros sitios.