Description
A vulnerability was identified in deepseek-ai deepseek-harness up to 0.1.1-rc.2. This affects the function agent of the component dsh-workflow-worker-thread. The manipulation leads to enforcement of behavioral workflow. The attack is possible to be carried out remotely. The exploit is publicly available and might be used. The vendor was contacted early about this disclosure but did not respond in any way.
Published: 2026-10-12
Score: 5.1 Medium
EPSS: n/a
KEV: No
Impact: Unintended workflow execution
Action: Assess Impact
AI Analysis

Impact

A flaw in the agent component of Deepseek Harness allows a remote attacker to manipulate the system and force the execution of a behavioral workflow that the user did not intend. The exploitation results in the enforcement of a predefined workflow without the user’s authorization, potentially leading to disruption of service or unintended automation. The vulnerability is tied to CWE‑840 and CWE‑841, indicating improper resource management and control over workflow execution.

Affected Systems

The impact applies to deepseek-ai Deepseek‑Harness up through release 0.1.1‑rc.2. Any deployment of the dsh‑workflow‑worker‑thread component in those versions is susceptible to the described abuse.

Risk and Exploitability

The CVSS score of 5.1 classifies the flaw as a moderate‑safety risk, and the exploit is publicly available with no documented exploitation probability (EPSS not available). The vulnerability is not listed in CISA’s KEV catalog, but the attacker can perform the attack remotely by sending a crafted request to the agent. No official patch or vendor response has been issued, so the vulnerability may remain exploitable until a fix is released.

Generated by OpenCVE AI on October 12, 2026 at 04:26 UTC.

Remediation

No solution or workaround provided in the CVE record.

OpenCVE Recommended Actions

  • Limit external network access to the dsh‑workflow‑worker‑thread service so that only trusted internal hosts can interact with it.
  • Enforce strict input validation on any commands or configuration changes received by the agent to prevent unauthorized workflow triggers.
  • Apply access controls or authentication mechanisms to restrict which users or services can invoke the agent’s workflow functions.

Generated by OpenCVE AI on October 12, 2026 at 04:26 UTC.

Tracking

Sign in to view the affected projects.

Advisories

No advisories yet.

History

Mon, 12 Oct 2026 03:00:00 +0000

Type Values Removed Values Added
Description A vulnerability was identified in deepseek-ai deepseek-harness up to 0.1.1-rc.2. This affects the function agent of the component dsh-workflow-worker-thread. The manipulation leads to enforcement of behavioral workflow. The attack is possible to be carried out remotely. The exploit is publicly available and might be used. The vendor was contacted early about this disclosure but did not respond in any way.
Title deepseek-ai deepseek-harness dsh-workflow-worker-thread agent behavioral workflow
First Time appeared Deepseek-ai
Deepseek-ai deepseek-harness
Weaknesses CWE-840
CWE-841
CPEs cpe:2.3:a:deepseek-ai:deepseek-harness:*:*:*:*:*:*:*:*
Vendors & Products Deepseek-ai
Deepseek-ai deepseek-harness
References
Metrics cvssV2_0

{'score': 6.5, 'vector': 'AV:N/AC:L/Au:S/C:P/I:P/A:P/E:POC/RL:ND/RC:UR'}

cvssV3_0

{'score': 5.5, 'vector': 'CVSS:3.0/AV:N/AC:L/PR:L/UI:R/S:U/C:L/I:L/A:L/E:P/RL:X/RC:R'}

cvssV3_1

{'score': 5.5, 'vector': 'CVSS:3.1/AV:N/AC:L/PR:L/UI:R/S:U/C:L/I:L/A:L/E:P/RL:X/RC:R'}

cvssV4_0

{'score': 5.1, 'vector': 'CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:P/VC:L/VI:L/VA:L/SC:N/SI:N/SA:N/E:P'}

Subscriptions

Deepseek-ai Deepseek-harness
cve-icon MITRE

Status: PUBLISHED

Assigner: VulDB

Published:

Updated: 2026-10-12T02:15:31.636Z

Reserved: 2026-10-11T08:45:05.631Z

Link: CVE-2026-108789

cve-icon Vulnrichment

No data.

cve-icon NVD

Status : Received

Published: 2026-10-12T03:17:17.820

Modified: 2026-10-12T03:17:17.820

Link: CVE-2026-108789

cve-icon Redhat

No data.

cve-icon OpenCVE Enrichment

Updated: 2026-10-12T04:30:06Z

Weaknesses