Skip to content
CVE Ledger
  • Newest CVEsPage
  • Recently added to KEVPage
  • Rising exploitation (EPSS)Page
  • Weekly digestPage
  • StatisticsPage
  • WatchlistPage
  • API docsPage
  • MCP serverPage
  • AboutPage
  • Data sourcesPage
  • Data statusPage
  • ChangelogPage
  • How it's builtPage
  • Support the projectPage

14 results

Back

CVE-2026-54236vLLM: incomplete CVE-2026-22778 fix leaks PIL repr addresses via Anthropic router

Description

vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.23.1rc0, the fix for CVE-2026-22778, which introduced a sanitize_message helper that strips object-repr memory addresses from error messages before they reach the client, is incomplete: several response paths echo str(exc) directly to clients without calling sanitize_message. The unsanitized sites include the Anthropic API router in vllm/entrypoints/anthropic/api_router.py (the POST /v1/messages and POST /v1/messages/count_tokens handlers), the Server-Sent Events streaming converter in vllm/entrypoints/anthropic/serving.py, and the realtime speech-to-text WebSocket in vllm/entrypoints/speech_to_text/realtime/connection.py. These paths catch the exception inside the route coroutine and construct the JSONResponse themselves, bypassing the sanitizing global FastAPI exception handler, and WebSocket frames do not traverse that handler chain at all. Using the same primitive as the parent issue, an unauthenticated attacker can send malformed image bytes through the Anthropic Messages API image content parts so that PIL.Image.open raises an UnidentifiedImageError whose message contains the BytesIO object repr, leaking the heap memory address verbatim in the error.message field of the response body. This vulnerability is fixed in 0.23.1rc0.

Public exploit

An exploit or proof of concept is public: anyone can try it against an unpatched system. The public collections list a Nuclei template for it. A reference of the record is tagged as an exploit. CISA's SSVC rates its exploitation: a proof of concept is public.

Show the exploits2

Nuclei templates1 template

References tagged exploit1 reference

Modules, templates and exploits as listed by the Metasploit Framework (BSD-3-Clause), ProjectDiscovery's Nuclei templates (MIT) and the Exploit Database (GPLv2), read once a day. Where the data comes from

Timeline

  1. ID reserved

    By GitHub_M

  2. · 10 days later

    Published

  3. · 1 day later

    Record updated

Changes after publication (a new score, a public exploit, EPSS crossing 1, 10 or 50 %, leaving KEV) are kept since November 2026; earlier ones are not known.

Scores

CVSS scores of CVE-2026-54236, by who scored it
Scored byCVSSScoreVector
GitHub_MCNA (assigner)3.1Medium5.3CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N
NVDNVD · Analyzed3.1Medium5.3CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:N/A:N
EPSS (exploitation in 30 days)
0.93% · top 39% of all CVEs (16 Nov 2026)
CISA SSVC
Exploitation: poc · Automatable: yes · Technical impact: partial
Loading the EPSS history…

The list uses the assigner's score when there is one, then CISA's, then NVD's: Medium 5.3.

Affected products

Affected packages

From reviewed advisories of the GitHub Advisory Database.

References

References3

Sources

The same CVE in the primary databases and aggregators, to check against.

Badge for a README or a ticket

CVE-2026-54236 badge
[![CVE-2026-54236](https://cveledger.app/badge/CVE-2026-54236.svg)](https://cveledger.app/cve/CVE-2026-54236)

Record last updated .