dbs-computer-use

Fail

Audited by Snyk on Nov 17, 2026

Risk Level: CRITICAL
Full Analysis

CRITICAL E004: Prompt injection detected in skill instructions.

  • Potential prompt injection detected (medium risk: 0.30). The skill is openly a computer-use tool with consent gates, but it sits in a risky middle ground: it nudges the user toward granting Cua access to an existing logged-in browser profile, a protection the skill itself calls account-risky, without malicious intent.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

  • Third-party content exposure detected (low risk: 0.10). The required workflow reads arbitrary web pages and desktop app content that the agent navigates to on the user's request (browser extraction and accessibility-tree reads), so outsiders can publish poison but cannot force delivery.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

  • Potentially malicious external URL detected (medium risk: 0.30). The skill's operational dependency is remote install scripts from cua.ai and opencode.ai, piped straight into a shell or iex during setup; cua.ai and opencode.ai match the Cua and OpenCode projects, but nothing enforces a signature or digest, so I rate it reputable-unofficial (medium) rather than strong.

MEDIUM W013: Attempt to modify system services in skill instructions.

  • Attempt to modify system services in skill instructions detected (low risk: 0.35). The skill mainly guides installing well-known computer-use tools with explicit user consent and has the user perform the system permission grants, but it also involves running downloaded official install scripts and enabling accessibility and screen-recording permissions, which are borderline state changes.

Issues (4)

E004
CRITICAL

Prompt injection detected in skill instructions.

W011
MEDIUM

Third-party content exposure detected (indirect prompt injection risk).

W012
MEDIUM

Unverifiable external dependency detected (runtime URL that controls agent).

W013
MEDIUM

Attempt to modify system services in skill instructions.

Audit Metadata
Risk Level
CRITICAL
Analyzed
Nov 17, 2026, 07:15 AM
Issues
4