Back to skill

Security audit

Standalone Setup

Security checks for vulnerabilities and agentic risk

Overview

This skill is a plausible VPS store installer, but it asks users to run an unverified remote installer as root and enables persistent automatic updates.

Review carefully before installing. Only run this on a disposable or well-backed-up VPS if you trust Mobazha's installer source, and prefer a pinned, checksum- or signature-verified installer. Use a domain with HTTPS or a secure SSH tunnel for first admin setup, and check how to disable or audit the hourly updater before using it for real store or payment data.

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • System PersistenceInstalls backdoors, hooks, services, or scheduled tasks that survive the run
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (3)

T03 · Remote Payload Retrieval and Execution

Error
Location
SKILL.md:43
Finding

Unpinned Remote Installer Is Executed with Root Privileges

Content
View full analysis
Remediation
View remediation
/standalone.sh echo " /tmp/mobazha-installer.sh" | sha256sum --check - sudo bash /tmp/mobazha-installer.sh rm -f /tmp/mobazha-installer.sh ``` The pinned version and checksum must come from a trusted, independently protected release channel. ]]>

T06 · System Persistence

Error
Location
SKILL.md:75
Finding

Installer Establishes a Persistent Hourly Auto-Update Mechanism

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:88
Finding

Initial Administrative Onboarding Is Offered over Plaintext HTTP

Content
View full analysis
/admin` On first visit, a **Setup Wizard** appears with 4 steps: 1. **Set admin password** — required before anything else ``` ### Technical Analysis The domainless deployment workflow tells the user to access the administrative onboarding interface over unencrypted HTTP. The first onboarding step requires creation of the administrator password. HTTP does not provide confidentiality, integrity, or server authentication. A network-positioned attacker can observe submitted credentials and session tokens, inject JavaScript or modified forms, redirect the user, or alter setup responses. The risk applies anywhere between the administrator and VPS, including local networks, public Wi-Fi, upstream providers, and compromised routing infrastructure. The fact that this is an initial setup workflow increases the severity because the first successful administrator registration may establish control over the entire store. ### Attack Path 1. The store is installed without a domain or TLS certificate. 2. The administrator opens `http:///admin` from a remote browser. 3. A network-positioned attacker intercepts or modifies the plaintext HTTP connection. 4. The attacker captures the submitted administrator password or session cookie, or injects code that sends those values elsewhere. 5. The attacker authenticates to the admin panel and takes control of the store. 6. Depending on application capabilities, the attacker modifies store configuration, accesses business data, changes payment settings, or uses management functionality to expand access. ### Impact Assessment Exploitation can expose the initial administrator password and a ...[truncated 315 chars]
Remediation
View remediation
Vulnerability Patterns
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Tool MisuseTool Parameter Abuse, Chaining Abuse, Unsafe Defaults
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Findings (36)

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

This is a textbook external-script-fetching vulnerability: curl retrieves code from a remote domain and immediately executes it. Because the script is not pinned, signed, hashed, or reviewed in the documented flow, any compromise of the delivery path becomes arbitrary code execution on the user's server.

Content

Scanner excerpt · SKILL.md (reported line 55)May include surrounding context.

Zero-config (most common):

bash
curl -sSL https://get.mobazha.org/standalone | sudo bash

With a pre-configured domain (enables auto-TLS via Let's Encrypt):

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

The pipe operator chains network retrieval directly into privileged execution, removing any inspection boundary between untrusted input and root-level command execution. This is dangerous because it compresses compromise and execution into a single step that is hard for users or tooling to audit.

Content

Scanner excerpt · SKILL.md (reported line 55)May include surrounding context.

Zero-config (most common):

bash
curl -sSL https://get.mobazha.org/standalone | sudo bash

With a pre-configured domain (enables auto-TLS via Let's Encrypt):

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

The domain-enabled command still fetches and executes untrusted remote content inline. The nearby warning suggesting users can review the script is helpful but insufficient because the actual prescribed command bypasses review and integrity verification entirely.

Content

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

With a pre-configured domain (enables auto-TLS via Let's Encrypt):

bash
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com

Privacy mode with Tor:

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

This chained command combines remote content ingestion with privileged shell execution while also accepting a domain parameter. The command structure itself is the issue: it treats untrusted network data as executable shell input without a trust-verification boundary.

Content

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

With a pre-configured domain (enables auto-TLS via Let's Encrypt):

bash
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com

Privacy mode with Tor:

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

Fetching and executing a Tor-install variant from a remote endpoint preserves the same supply-chain and delivery-path risks as the base installer. In a skill that requests SSH/root access to a VPS, this is especially dangerous because compromise yields full control of an internet-facing host.

Content

Scanner excerpt · SKILL.md (reported line 67)May include surrounding context.

Privacy mode with Tor:

bash
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --overlay tor

Using testnets (for trying things out without real money):

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

The Tor overlay example uses the same dangerous chaining pattern, where a remote response stream is fed straight to sudo bash. Because the skill is meant for autonomous or semi-automated setup on user infrastructure, this pattern materially increases the risk of supply-chain compromise turning into full-host takeover.

Content

Scanner excerpt · SKILL.md (reported line 67)May include surrounding context.

Privacy mode with Tor:

bash
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --overlay tor

Using testnets (for trying things out without real money):

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

The testnet example still performs immediate execution of remote code, so the environment label does not materially lower the vulnerability. Attackers can exploit development or test infrastructure just as effectively, especially when it is exposed on a public VPS.

Content

Scanner excerpt · SKILL.md (reported line 73)May include surrounding context.

Using testnets (for trying things out without real money):

bash
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --testnet

Flags can be combined:

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

This testnet installation still abuses command chaining by transforming downloaded bytes into root-executed commands in one shell pipeline. The 'testnet' context does not mitigate host-level compromise or subsequent abuse of the VPS.

Content

Scanner excerpt · SKILL.md (reported line 73)May include surrounding context.

Using testnets (for trying things out without real money):

bash
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --testnet

Flags can be combined:

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

This combined-flags example normalizes a one-liner that both fetches and executes arbitrary remote code under elevated privileges. The convenience-focused presentation increases the chance users will run it without scrutiny, making the skill context more dangerous.

Content

Scanner excerpt · SKILL.md (reported line 79)May include surrounding context.

Flags can be combined:

bash
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com --overlay tor --testnet

Step 3: What Happens During Installation

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

This final combined example is especially risky because it packages multiple options into a convenience one-liner that preserves the insecure retrieval-to-root-execution chain. In practice, such examples are likely to be copied verbatim, increasing the probability of unsafe deployment.

Content

Scanner excerpt · SKILL.md (reported line 79)May include surrounding context.

Flags can be combined:

bash
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com --overlay tor --testnet

Step 3: What Happens During Installation

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

The document's primary installation pattern fetches a script from an external URL and executes it immediately, preventing inspection and bypassing integrity controls. In a server-deployment skill, this is particularly dangerous because a single upstream compromise can yield arbitrary root code execution on production VPS instances.

Content

Scanner excerpt · references/install-flags.md (reported line 3)May include surrounding context.

md
# Standalone Installer Flags Reference

Source: `curl -sSL https://get.mobazha.org/standalone | sudo bash [options]`

## Available Flags

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

The shell pipeline chains network retrieval directly into privileged execution, eliminating any opportunity for the user or tooling to validate the content in between. This chaining is dangerous because it amplifies any upstream compromise into immediate root-level code execution on the target host.

Content

Scanner excerpt · references/install-flags.md (reported line 3)May include surrounding context.

md
# Standalone Installer Flags Reference

Source: `curl -sSL https://get.mobazha.org/standalone | sudo bash [options]`

## Available Flags

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

This basic install example normalizes fetching executable content from the internet and running it instantly. Because the target is a self-hosted store on a Linux VPS, exploitation could lead to full takeover of the storefront, data theft, malware installation, and lateral movement from the server.

Content

Scanner excerpt · references/install-flags.md (reported line 26)May include surrounding context.

bash
# Basic install — HTTP on IP
curl -sSL https://get.mobazha.org/standalone | sudo bash

# With domain — HTTPS auto-TLS
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

This command uses a pipeline to stream remote script content straight into sudo bash, which is a classic chaining-abuse pattern. The danger is heightened in deployment documentation because users are likely to copy-paste it exactly, giving attackers a low-friction path to compromise servers if the source is tampered with.

Content

Scanner excerpt · references/install-flags.md (reported line 26)May include surrounding context.

bash
# Basic install — HTTP on IP
curl -sSL https://get.mobazha.org/standalone | sudo bash

# With domain — HTTPS auto-TLS
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

The domain-enabled example still executes unverified remote code from an external endpoint, with no integrity validation or provenance check. Since domain/TLS configuration is part of the install path, a malicious installer could also abuse certificate, DNS, or reverse-proxy setup to intercept traffic or persist control.

Content

Scanner excerpt · references/install-flags.md (reported line 29)May include surrounding context.

md
curl -sSL https://get.mobazha.org/standalone | sudo bash

# With domain — HTTPS auto-TLS
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com

# Privacy mode
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --overlay tor

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

The pipeline on this line still streams untrusted external content directly into a privileged shell. The presence of configuration flags does not mitigate the risk; instead, it encourages repeated reuse of the unsafe pattern across different deployment scenarios.

Content

Scanner excerpt · references/install-flags.md (reported line 29)May include surrounding context.

md
curl -sSL https://get.mobazha.org/standalone | sudo bash

# With domain — HTTPS auto-TLS
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com

# Privacy mode
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --overlay tor

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

This privacy mode example repeats the external-script execution pattern, exposing users to arbitrary code from the remote source. The context of configuring overlay networking can make host compromise harder to detect and easier for an attacker to hide once the installer is abused.

Content

Scanner excerpt · references/install-flags.md (reported line 32)May include surrounding context.

md
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com

# Privacy mode
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --overlay tor

# Combined
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com --overlay tor --testnet

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

This example continues to chain fetch and execution with root privileges, leaving no control point for integrity verification. Because it is part of a self-hosted infrastructure setup, a successful exploit can yield persistent compromise of a public-facing VPS and any associated service credentials.

Content

Scanner excerpt · references/install-flags.md (reported line 32)May include surrounding context.

md
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com

# Privacy mode
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --overlay tor

# Combined
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com --overlay tor --testnet

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

The combined example continues the unsafe pattern of executing a remote script directly from an external URL. Because this installation configures a public-facing commerce service, compromise can affect availability, customer data, wallet/payment integrations, and long-term server integrity.

Content

Scanner excerpt · references/install-flags.md (reported line 35)May include surrounding context.

md
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --overlay tor

# Combined
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com --overlay tor --testnet

# Custom install directory
INSTALL_DIR=/srv/mobazha curl -sSL https://get.mobazha.org/standalone | sudo bash

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

The combined install command preserves the chaining-abuse issue by piping remote content straight into a root shell. This is especially risky in a multi-option deployment example because it presents the unsafe pattern as the standard method for more complex, likely production-grade installs.

Content

Scanner excerpt · references/install-flags.md (reported line 35)May include surrounding context.

md
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --overlay tor

# Combined
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com --overlay tor --testnet

# Custom install directory
INSTALL_DIR=/srv/mobazha curl -sSL https://get.mobazha.org/standalone | sudo bash

External Script Fetching

High
Category
Supply Chain
Confidence
99% confidence
Finding

This example combines an environment variable with immediate execution of a remotely fetched script, preserving the same supply-chain risk. If the external source is compromised, attackers gain arbitrary execution during installation and may exploit environment-driven behavior to place files or persistence in attacker-chosen locations.

Content

Scanner excerpt · references/install-flags.md (reported line 38)May include surrounding context.

curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com --overlay tor --testnet

Custom install directory

INSTALL_DIR=/srv/mobazha curl -sSL https://get.mobazha.org/standalone | sudo bash

text

## Post-Install Locations

Chaining Abuse

High
Category
Tool Misuse
Confidence
98% confidence
Finding

This line combines environment-variable injection with a pipeline from remote fetch to privileged execution, increasing opacity and risk. The chained form makes it easy for attackers controlling the remote script to exploit both immediate execution and installer parameters for deeper host compromise.

Content

Scanner excerpt · references/install-flags.md (reported line 38)May include surrounding context.

curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com --overlay tor --testnet

Custom install directory

INSTALL_DIR=/srv/mobazha curl -sSL https://get.mobazha.org/standalone | sudo bash

text

## Post-Install Locations

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
70% confidence
Finding

Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

Content

Scanner excerpt · SKILL.md (reported line 18)May include surrounding context.

md
## Prerequisites

- A Linux VPS (Ubuntu 22.04+ or Debian 12+) with 2+ CPU cores, 2+ GB RAM, 20+ GB SSD
- Root or sudo access
- (Optional) A domain name pointed to the VPS IP

### Recommended VPS Providers

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
98% confidence
Finding

The skill instructs users to download a remote script and execute it immediately with sudo via bash. This creates a direct remote-code-execution path as root if the upstream server, DNS, TLS trust chain, or script content is compromised, and the damage is amplified because the command is designed for unattended execution.

Content

Scanner excerpt · SKILL.md (reported line 55)May include surrounding context.

Zero-config (most common):

bash
curl -sSL https://get.mobazha.org/standalone | sudo bash

With a pre-configured domain (enables auto-TLS via Let's Encrypt):

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
98% confidence
Finding

This variant still executes a remote script as root, only with a domain argument appended. User-supplied parameters do not reduce the underlying risk that arbitrary code from an external host is being run with full system privileges.

Content

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

With a pre-configured domain (enables auto-TLS via Let's Encrypt):

bash
curl -sSL https://get.mobazha.org/standalone | sudo bash -s -- --domain shop.example.com

Privacy mode with Tor:

Static analysis

No suspicious patterns detected.