A vulnerability was found in bentoml OpenLLM 0.6.30. This affects the function async_run_command of the file src/openllm/common.py of the component Model Repository Directory Name Handler. Performing a manipulation of the argument cmd results in command injection. Attacking locally is a requirement. The exploit has been made public and could be used. The project was informed of the problem early through an issue report but has not responded yet.
CVE-2026-15035
Score 5.3 from GitHub Security Advisory (severity: LOW) published 2026-07-08. NVD baseline CVSS 7.8; sources differ by 2.5.
- High severity, but no confirmed exploitation yet
No vendor fix yet — apply a workaround or compensating control (WAF / firewall / segmentation) and watch for a patch.
- CVSS v3
- 7.8
- EG Score
- 5.3(medium)
- EPSS
- 68.0%
- KEV
- Not listed
Published
July 8, 2026
Last Modified
July 9, 2026
Advisory Details (3)
Auto-updated Jul 9, 2026fix(security): use exec instead of shell in async_run_command (CWE-78)
Upstream fix in progress (PR #1235 · bentoml/OpenLLM, state=open)
https://github.com/bentoml/OpenLLM/pull/1235[Security] Shell command injection via malicious model repository directory names · Issue #1229 · bentoml/OpenLLM · GitHub
https://github.com/bentoml/OpenLLM/issues/1229GitHub - bentoml/OpenLLM: Run any open-source LLMs, such as DeepSeek and Llama, as OpenAI compatible API endpoint in the cloud. · GitHub
https://github.com/bentoml/OpenLLM/Vendor Advisories for CVE-2026-15035(1)
These vendors published their own advisory mentioning this CVE — often with vendor-specific remediation steps + affected product lists not in NVD.
Affected Packages
(1 across 1 ecosystem)
PyPI(1)
| Package | Vulnerable range | Fixed in | Dependents |
|---|---|---|---|
| bentoml | 0.0.1 ... 0.6.3 (41 versions) | — | — |
Weakness Classification(3)
MITRE Common Weakness Enumeration — the root-cause categories this CVE belongs to.
Data Freshness Timeline
(refreshed 23× in last 7d / 23× in last 30d)
Each row is a source pipeline that fetched or updated this CVE on that date, with what changed. 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. Most recent first.
- 2026-07-14 02:06 UTCEG score recompute
- 2026-07-14 02:06 UTCGHSA enrichment
- 2026-07-13 22:30 UTCEPSS rescore
- 2026-07-13 06:12 UTCEPSS rescore
- 2026-07-13 04:07 UTCEG score recompute
- 2026-07-13 04:06 UTCGHSA enrichment
- 2026-07-12 06:03 UTCEG score recompute
- 2026-07-12 06:02 UTCGHSA enrichment
- 2026-07-12 05:46 UTCEPSS rescore
- 2026-07-11 08:27 UTCEPSS rescore
- 2026-07-11 08:03 UTCEG score recompute
- 2026-07-11 08:03 UTCGHSA enrichment
- 2026-07-10 10:04 UTCEG score recompute
- 2026-07-10 10:04 UTCGHSA enrichment
- 2026-07-09 19:10 UTCEPSS rescore
- 2026-07-09 17:00 UTCNVD updateCVSS v3 → 7.8 · severity → HIGH
- 2026-07-09 12:05 UTCEG score recompute
- 2026-07-09 12:05 UTCGHSA enrichment
- 2026-07-08 15:13 UTCNVD updateCVSS v4 → 1.9
- 2026-07-08 15:00 UTCMITRE cvelistV5CVSS v4 → 4.8
- 2026-07-08 14:36 UTCNVD updateCVSS v4 → 1.9
- 2026-07-08 14:06 UTCEG score recompute
- 2026-07-08 14:05 UTCMITRE cvelistV5first tracked
Related CVEs(same CWE)
Same CWE
10 shownCWE-78 · CWE-74 · CWE-77
- CVE-2009-20011EG 10.0CRITICAL
- CVE-2011-10017EG 10.0CRITICAL
- CVE-2011-10026EG 9.8CRITICAL
- CVE-2011-2195EG 9.8CRITICAL
- CVE-2011-2717EG 9.8CRITICAL
- CVE-2011-2523EG 9.8EPSS 100%CRITICAL
- CVE-2005-3056EG 9.8CRITICAL
- CVE-2010-5330EG 9.8 KEVEPSS 98%CRITICAL
- CVE-2009-5156EG 9.8EPSS 95%CRITICAL
- CVE-2007-3010EG 9.8 KEVEPSS 100%CRITICAL
Frequently asked(5)
What is CVE-2026-15035?
When was CVE-2026-15035 disclosed?
Is CVE-2026-15035 actively exploited?
What is the CVSS score of CVE-2026-15035?
How do I remediate CVE-2026-15035?
Dependency Blast Radius
See which npm, PyPI, Go, and Maven packages are affected by CVE-2026-15035
Is Your Infrastructure Affected by CVE-2026-15035?
EchelonGraph automatically scans your cloud infrastructure and maps CVE exposure using blast radius analysis.