LiteLLM is a proxy server (AI Gateway) to call LLM APIs in OpenAI (or native) format. From version 1.74.2 to before version 1.83.7, two endpoints used to preview an MCP server before saving it — POST /mcp-rest/test/connection and POST /mcp-rest/test/tools/list — accepted a full server configuration in the request body, including the command, args, and env fields used by the stdio transport. When called with a stdio configuration, the endpoints attempted to connect, which spawned the supplied command as a subprocess on the proxy host with the privileges of the proxy process. The endpoints were gated only by a valid proxy API key, with no role check. Any authenticated user — including holders of low-privilege internal-user keys — could therefore run arbitrary commands on the host. This issue has been patched in version 1.83.7.
CVE-2026-42271
“▲” beside the score: it is higher than NVD's CVSS score.
Score elevated to 9.0 because this CVE is listed on the CISA Known Exploited Vulnerabilities catalog (added 2026-06-08), indicating real-world exploitation has been confirmed by US federal agencies. NVD baseline CVSS 8.8 retained for reference. Confidence: HIGH.
- Actively exploited in the wild (CISA-KEV) — the reason this is sooner than the SSVC decision on this page at low or medium mission impact
A fix is available — apply it.
- CVSS v3
- 8.8
- EG Score
- 9.0CRITICALhigh confidence
- EG Risk
- 81EG Risk 81/100
EG Risk is EchelonGraph's 0–100 priority score: it fuses intrinsic severity with real-world exploitation and automatability so you can rank equal-severity CVEs and fix the most dangerous first. Higher = act sooner. Distinct from the 0–10 EG Score (severity).
How it’s computed (value × EchelonGraph’s weight)Severity90% × 45%The EG Score, 9.0 of 10.Exploitation100% × 40%Fixed at 100% for a CISA KEV listing: EchelonGraph's rule.Automatability0% × 15%CISA's SSVC Automatable value: no (Vulnrichment).SSVC (CISA's decision table, applied by EchelonGraph): Track at low, Attend at medium and Act at high mission impact.Exploitation active (CISA Vulnrichment) · Automatable no (CISA Vulnrichment) · Technical impact total (CISA Vulnrichment). All three are CISA's SSVC values (Vulnrichment). EchelonGraph applied CISA's decision table to them; CISA publishes SSVC inputs and the table, not a decision for each CVE.Action: A fix is available. Apply it within your standard update timelines at low, sooner than that at medium and as soon as possible at high mission impact. - EPSS PROB
- 93%
- EPSS %ILE
- 99th
- KEV
- ⚠ Exploited
SSVC (CISA's decision table, applied by EchelonGraph)Track at low, Attend at medium and Act at high mission impact.
A fix is available. Apply it within your standard update timelines at low, sooner than that at medium and as soon as possible at high mission impact.
Exploitation active (CISA Vulnrichment) · Automatable no (CISA Vulnrichment) · Technical impact total (CISA Vulnrichment). All three are CISA's SSVC values (Vulnrichment). EchelonGraph applied CISA's decision table to them; CISA publishes SSVC inputs and the table, not a decision for each CVE. Mission impact is CISA's Mission & Well-being decision point, and only you can judge it. CISA's table makes it high when Mission Prevalence is Essential — the vulnerable component "directly provides capabilities that constitute at least one MEF for at least one entity" (MEF: mission essential function) — or when Public Well-Being Impact is Irreversible: "multiple fatalities are likely", the cyber-physical system "is likely lost or destroyed", "extreme or serious externalities" are imposed on other parties, or social systems such as elections or the financial grid "are destabilized and potentially collapse". CISA's decision table
BOD 26-04 (CISA's remediation timeline, Table 1)3 days & forensic triage if the affected system is publicly exposed; 14 days if it is not.
In CISA KEV yes (CISA's KEV catalog) · Automatable no (CISA Vulnrichment) · Technical impact total (CISA Vulnrichment). Whether your system is publicly exposed is yours to answer. BOD 26-04 binds US Federal Civilian Executive Branch agencies; for anyone else it is CISA's published timeline, a reference rather than an obligation. The directive · how EchelonGraph applies it
CISA required action“Apply mitigations per vendor instructions, follow applicable BOD 22-01 guidance for cloud services, or discontinue use of the product if mitigations are unavailable.”
Quoted verbatim from CISA's Known Exploited Vulnerabilities catalog. CISA directs it at US federal civilian executive branch agencies under Binding Operational Directive (BOD) 26-04, which replaced BOD 22-01 on 10 June 2026; it is not EchelonGraph's advice. CISA's notes: https://github.com/BerriAI/litellm/security/advisories/GHSA-v4p8-mg3p-g94g · https://github.com/BerriAI/litellm/releases/tag/v1.83.7-stable · https://nvd.nist.gov/vuln/detail/CVE-2026-42271
Published
May 8, 2026
Last Modified
July 15, 2026
Advisory Details (9)
Auto-updated Oct 12, 2026Known Exploited Vulnerabilities Catalog | CISA
Known Exploited Vulnerabilities Catalog | CISA. Listed in CISA Known Exploited Vulnerabilities catalog.
https://www.cisa.gov/known-exploited-vulnerabilities-catalog?field_cve=CVE-2026-422712467924 – (CVE-2026-42271) CVE-2026-42271 litellm: LiteLLM: Authenticated command execution via MCP stdio test endpoints
https://bugzilla.redhat.com/show_bug.cgi?id=2467924CVE-2026-42271 - Red Hat Customer Portal
Affected: Red Hat Enterprise Linux 8 or OpenShift Container Platform 4, is affected by this vulnerability and a fix may be released to a
https://access.redhat.com/security/cve/CVE-2026-42271Authenticated command execution via MCP stdio test endpoints · Advisory · BerriAI/litellm · GitHub
https://github.com/BerriAI/litellm/security/advisories/GHSA-v4p8-mg3p-g94gv1.83.7-stable
Patch available: BerriAI/litellm v1.83.7-stable
https://github.com/BerriAI/litellm/releases/tag/v1.83.7-stableVendor Advisories for CVE-2026-42271(4)
These vendors published their own advisory mentioning this CVE — often with vendor-specific remediation steps + affected product lists not in NVD.
- RHSA-2026:30056Red Hat Product SecurityHigh
Red Hat Security Advisory: RHOAI 3.3.4 - Red Hat OpenShift AI
- RHSA-2026:28960Red Hat Product SecurityHigh
Red Hat Security Advisory: RHOAI 2.25.8 - Red Hat OpenShift AI
- RHSA-2026:27784Red Hat Product SecurityHigh
Red Hat Security Advisory: RHOAI 3.4.1 - Red Hat OpenShift AI
- GHSA-v4p8-mg3p-g94gGitHub Security AdvisoriesHigh
LiteLLM: Authenticated command execution via MCP stdio test endpoints
Patch Availability(3)
| Vendor / Ecosystem | Fix (by version range) / Patch | Released | Source |
|---|---|---|---|
| redhat | rhoai/odh-llama-stack-core-rhel9:1782310008 | 2026-06-25 | redhat |
| redhat | rhoai/odh-llama-stack-core-rhel9:1781826406 | 2026-06-24 | redhat |
| redhat | rhoai/odh-trustyai-garak-lls-provider-dsp-rhel9:1781622627 | 2026-06-22 | redhat |
Patches are aggregated from vendor advisories (Red Hat, Microsoft, Cisco, GitHub) and package ecosystems (OSV, GHSA). Multiple rows for the same upstream release have been deduplicated.
Affected Packages
(1 across 1 ecosystem)
PyPI(1)
| Package | Vulnerable range | Fix by version range | Dependents |
|---|---|---|---|
| litellm | 1.74.12 ... 1.83.6 (107 versions) |
| — |
Weakness Classification(2)
MITRE Common Weakness Enumeration — the root-cause categories this CVE belongs to.
Data Freshness Timeline
(refreshed 30× in last 7d / 199× in last 30d)
Each row is a time one of our pipelines fetched, updated or re-scored this CVE. For example, "NVD update" means NVD published or revised its analysis for this CVE; "MITRE cvelistV5" means we ingested or refreshed it from the CNA feed. An arrow marks a change the record states ("CVSS v3 → 9.1"); EPSS and GHSA rows show the values recorded at that time, not what changed (EPSS values are rounded; hover one for the exact figures). Most recent first.
Showing the most recent 100 of 871 refreshes on record for this CVE.
A repeat, a row identical to one the same pipeline wrote less than 10 seconds earlier, is not a separate refresh and is not counted. None are among the latest 100 rows. The 771 oldest rows on record are not shown and were not checked, so the counts can still include repeats.
- 2026-10-16 11:49 UTCEG score recompute
- 2026-10-16 11:49 UTCVendor advisory
- 2026-10-15 12:34 UTCEG score recompute
- 2026-10-15 12:34 UTCVendor advisory
- 2026-10-14 16:33 UTCEPSS rescoreEPSS 93% · 99th percentile
- 2026-10-14 13:19 UTCEG score recompute
- 2026-10-14 13:19 UTCVendor advisory
- 2026-10-13 16:34 UTCEPSS rescoreEPSS 93% · 99th percentile
- 2026-10-13 14:04 UTCEG score recompute
- 2026-10-13 14:04 UTCVendor advisory
- 2026-10-12 16:34 UTCEPSS rescoreEPSS 93% · 99th percentile
- 2026-10-12 14:49 UTCEG score recompute
- 2026-10-12 14:49 UTCVendor advisory
- 2026-10-12 06:15 UTCVendor advisory
- 2026-10-12 02:15 UTCVendor advisory
- 2026-10-11 22:15 UTCVendor advisory
- 2026-10-11 19:49 UTCEG score recompute
- 2026-10-11 18:16 UTCVendor advisory
- 2026-10-11 17:03 UTCEG score recompute
- 2026-10-11 14:16 UTCEG score recompute
- 2026-10-11 14:16 UTCVendor advisory
- 2026-10-11 13:48 UTCEPSS rescoreEPSS 93% · 99th percentile
- 2026-10-11 10:19 UTCVendor advisory
- 2026-10-11 06:21 UTCVendor advisory
- 2026-10-11 02:21 UTCVendor advisory
Show 75 moreShow fewer
- 2026-10-10 22:21 UTCVendor advisory
- 2026-10-10 18:23 UTCVendor advisory
- 2026-10-10 14:25 UTCVendor advisory
- 2026-10-10 10:27 UTCVendor advisory
- 2026-10-10 06:29 UTCVendor advisory
- 2026-10-10 02:31 UTCVendor advisory
- 2026-10-09 22:33 UTCVendor advisory
- 2026-10-09 18:34 UTCVendor advisory
- 2026-10-09 14:36 UTCVendor advisory
- 2026-10-09 10:36 UTCVendor advisory
- 2026-10-09 06:37 UTCVendor advisory
- 2026-10-09 02:39 UTCVendor advisory
- 2026-10-08 22:41 UTCVendor advisory
- 2026-10-08 18:44 UTCVendor advisory
- 2026-10-08 14:46 UTCVendor advisory
- 2026-10-08 10:47 UTCVendor advisory
- 2026-10-08 06:49 UTCVendor advisory
- 2026-10-08 02:51 UTCVendor advisory
- 2026-10-07 22:54 UTCVendor advisory
- 2026-10-07 18:55 UTCVendor advisory
- 2026-10-07 14:58 UTCVendor advisory
- 2026-10-07 11:00 UTCVendor advisory
- 2026-10-07 07:03 UTCVendor advisory
- 2026-10-07 03:05 UTCVendor advisory
- 2026-10-06 23:07 UTCVendor advisory
- 2026-10-06 19:09 UTCEG score recompute
- 2026-10-06 19:09 UTCVendor advisory
- 2026-10-06 15:17 UTCEPSS rescoreEPSS 93% · 99th percentile
- 2026-10-06 15:12 UTCVendor advisory
- 2026-10-06 11:12 UTCVendor advisory
- 2026-10-06 07:15 UTCVendor advisory
- 2026-10-06 03:17 UTCVendor advisory
- 2026-10-06 03:17 UTCVendor advisory
- 2026-10-06 03:17 UTCVendor advisory
- 2026-10-05 23:18 UTCVendor advisory
- 2026-10-05 19:20 UTCVendor advisory
- 2026-10-05 15:21 UTCVendor advisory
- 2026-10-05 11:23 UTCVendor advisory
- 2026-10-05 07:26 UTCVendor advisory
- 2026-10-05 03:28 UTCVendor advisory
- 2026-10-04 23:31 UTCVendor advisory
- 2026-10-04 19:33 UTCVendor advisory
- 2026-10-04 15:36 UTCVendor advisory
- 2026-10-04 11:38 UTCVendor advisory
- 2026-10-04 07:41 UTCVendor advisory
- 2026-10-04 03:43 UTCVendor advisory
- 2026-10-03 23:46 UTCVendor advisory
- 2026-10-03 19:48 UTCVendor advisory
- 2026-10-03 15:50 UTCVendor advisory
- 2026-10-03 11:53 UTCVendor advisory
- 2026-10-03 07:55 UTCVendor advisory
- 2026-10-03 03:55 UTCVendor advisory
- 2026-10-02 23:55 UTCVendor advisory
- 2026-10-02 19:54 UTCVendor advisory
- 2026-10-02 15:56 UTCVendor advisory
- 2026-10-02 11:58 UTCVendor advisory
- 2026-10-02 07:57 UTCVendor advisory
- 2026-10-02 03:58 UTCVendor advisory
- 2026-10-02 00:00 UTCVendor advisory
- 2026-10-01 20:03 UTCVendor advisory
- 2026-10-01 16:06 UTCVendor advisory
- 2026-10-01 12:08 UTCVendor advisory
- 2026-10-01 08:10 UTCVendor advisory
- 2026-10-01 04:12 UTCVendor advisory
- 2026-10-01 00:15 UTCVendor advisory
- 2026-09-30 20:17 UTCVendor advisory
- 2026-09-30 16:20 UTCVendor advisory
- 2026-09-30 12:22 UTCVendor advisory
- 2026-09-30 08:24 UTCVendor advisory
- 2026-09-30 04:26 UTCVendor advisory
- 2026-09-30 00:28 UTCVendor advisory
- 2026-09-29 20:30 UTCVendor advisory
- 2026-09-29 16:32 UTCEG score recompute
- 2026-09-29 16:32 UTCVendor advisory
- 2026-09-29 12:32 UTCVendor advisory
Publicly available exploits
(3 references)Public exploit code is referenced for this CVE (1 Metasploit module, 1 GitHub PoC), as is 1 Nuclei detection template. A detection template checks whether a system is affected; it is not necessarily exploit code. Defenders should treat patch urgency accordingly.
- Metasploitexploit/multi/http/litellm_mcp_test_cmd_injection✓ verifiedPublished Sep 28, 2026
BerriAI LiteLLM Proxy MCP Test Endpoint Command Execution
Open source ↗ - GitHub PoCamnsecurity/CVE-2026-42271-LiteLLM-RCEPublished Jul 7, 2026
CVE-2026-42271 - LiteLLM AI Gateway MCP Command Injection RCE - PoC & Analysis | CVSS 8.8 | AMN SECURITY
Open source ↗ - Nucleihttp/cves/2026/CVE-2026-42271.yamlPublished Jun 3, 2026
LiteLLM - Command Injection
Open source ↗
Related CVEs(same product + same vendor + same CWE)
Same product
10 shownPyPI:litellm
Same vendor
3 shownredhat:RHSA-2026:28960 · redhat:RHSA-2026:30056
Frequently asked(6)
What is CVE-2026-42271?
When was the CVE record for CVE-2026-42271 published?
Is CVE-2026-42271 actively exploited?
What is the CVSS score of CVE-2026-42271?
Which products are affected by CVE-2026-42271?
How do I remediate CVE-2026-42271?
Dependency Blast Radius
See which npm, PyPI, Go, and Maven packages are affected by CVE-2026-42271
Is Your Infrastructure Affected by CVE-2026-42271?
EchelonGraph automatically scans your cloud infrastructure and maps CVE exposure using blast radius analysis.