Back to skill

Security audit

Dixa

Security checks for vulnerabilities and agentic risk

Overview

The Dixa features fit their purpose, but setup can execute downloaded code and make lasting agent changes without clearly requiring your approval.

Review the setup before installing. Prefer a reviewed, version-pinned CLI installation, and require approval before an ordinary Dixa request triggers software installation. Review automatic bundled-skill synchronization and grant only necessary Dixa access. OOMOL mediates customer-data access; remote installers and hosted connector internals remain unverified.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (4)

Description-Behavior Mismatch

High
Category
MCP Tool Poisoning
Confidence
99% confidence
Finding

The supplied code contains only a remote-script execution command and no Dixa search or read operations. Although installing a dependency can be a supporting implementation detail, this chunk establishes no relationship between the Oomol installer and the declared Dixa functionality. The downloaded script's contents and downstream effects are not provided and cannot be determined.

Content

No source excerpt is available for this finding.

External Script Fetching

High
Category
Supply Chain
Confidence
98% confidence
Finding

Piping a live HTTPS response directly into Bash executes unverified remote code with the invoking user's permissions. Compromise of the installer delivery infrastructure could expose local files and credentials or permit persistent changes; the legitimate dependency context and conditional fallback reduce exposure but do not establish code integrity.

Content

Scanner excerpt · SKILL.md (reported line 61)May include surrounding context.

The fallback setup executes remotely downloaded installation scripts without warning the user about system changes or remote-code execution.

Medium
Category
Output Handling
Confidence
94% confidence
Finding

The fallback directs installation of software without explicitly obtaining approval for executing remote code or changing the local environment. Although this is a legitimate dependency setup step and only occurs after a missing-command error, a request to read Dixa data does not by itself authorize local software installation.

Content

No source excerpt is available for this finding.

Autonomous Decision Making

Medium
Category
Excessive Agency
Confidence
80% confidence
Finding

Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.

Content

Scanner excerpt · SKILL.md (reported line 82)May include surrounding context.

md
- **`OOMOL_INSUFFICIENT_CREDIT`** — OOMOL billing stop. Recharge at `https://console.oomol.com/billing/token-recharge` before retrying.

- **HTTP 402 without `OOMOL_INSUFFICIENT_CREDIT`** — the third-party account itself has no active plan or credit. Report the provider error message as is; do not ask the user to recharge OOMOL.

## Resources