Connect your agent
AdvisoriesCriticalCheck your lockfilePluginHow it worksPricing Connect your agent

Advisory record · GHSA-72w5-pf8h-xfp4

DeepSeek TUI: task_create Insecure Defaults Enable RCE via Prompt Injection in Project Files

CVSS 9.6 CriticalCriticalEPSS 0.0039 · 33rd percentile

Critical, because:

  • CVSS score 9.6: 9.0 or more.

How we decide what’s critical

Published by the GitHub Advisory Database at 14 May 2026, 20:29 UTC. agentprime has not analysed this yet, so whether it affects any app is not known.

PackageEcosystemAffectedFixed in
deepseek-tuicrates.io<0.8.260.8.26
Identifiers
GHSA-72w5-pf8h-xfp4 · CVE-2026-45374
Weakness (CWE)
CWE-94
Tags
Published
14 May 2026
Updated
9 Jun 2026

Machine-written summary

Summary written by Claude Haiku from the source advisory. Not reviewed by a person.

DeepSeek TUI (deepseek-tui crate) versions before 0.8.26 are vulnerable. The task_create tool spawns sub-agents that default to allowing shell access and auto-approving actions, so a user who approves a task creation request may unknowingly grant unapproved shell execution. An attacker who gets a user to open a malicious project containing crafted instructions can cause the sub-agent to run commands on the user's machine. Upgrade to version 0.8.26 or later.

Written 8 Nov 2026.

Not analysed yet

We haven’t published our analysis of this Advisory yet. Once a person has reviewed it, this page will say what’s vulnerable, what has to be true for your app to be affected, and how to check.

Severity and exploitation

ScoreRatingCVSSVectorSource
9.6 Critical 3.1 CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:H/I:H/A:H CNA (GitHub_M)
  • CISA KEV Not listed.
  • EPSS 0.0039, the 33rd percentile, on 14 Nov 2026: FIRST’s estimate of the chance of exploitation in the next 30 days.

Timeline

  1. 31 Oct 2026, 15:18 UTC First seen by agentprime
  2. 31 Oct 2026, 15:18 UTC Confirmed by OSV

Sources

Facts on this page come from the GitHub Advisory Database (CC-BY 4.0), OSV and the other sources listed, as published. We don’t reproduce full advisory descriptions, which can contain exploit code.