Back to skill

Security audit

Okx Payments

Security checks for vulnerabilities and agentic risk

Overview

The skill has a clear payment-setup purpose, but it asks agents to run mutable remote install scripts and install unpinned agent skills, which needs user review.

Review the remote installer scripts and resolved skill/plugin source before installing. Prefer pinned releases or commit hashes with checksum or signature verification, and do not provide OKX API credentials or wallet-related values until you have verified the generated code and dependencies.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
Findings (3)

T03 · Remote Payload Retrieval and Execution

Error
Location
SKILL.md:29
Finding

Mutable Remote Installation Scripts Executed Directly by the Shell

Content
View full analysis
**Tip:** Detect the user's operating system first, then run the appropriate install command below. Do not run all commands — only the one matching the detected OS. ```bash # macOS / Linux curl -sSL https://raw.githubusercontent.com/okx/onchainos-skills/main/install.sh | sh # Windows PowerShell irm https://raw.githubusercontent.com/okx/onchainos-skills/main/install.ps1 | iex ``` ``` ### Technical Analysis Both installation commands retrieve scripts from the mutable `main` branch of an external GitHub repository and immediately execute them with the invoking user's privileges. The Linux/macOS command pipes the HTTP response directly to `sh`, while the Windows command passes it to `Invoke-Expression`. No immutable commit, release version, cryptographic checksum, or signature is specified. Consequently, the executable payload can change after this Skill has been audited. The security note following the commands recommends reviewing the scripts but does not technically enforce review, integrity verification, or user approval. This behavior is not necessary for the Skill's core function as an interactive setup guide. The Skill could instead provide verified download instructions or direct the user to reviewed installation documentation without automatically executing remote content. ### Attack Path 1. An attacker compromises the upstream repository, an authorized maintainer account, or the publication workflow. 2. The attacker replaces `install.sh` or `install.ps1` on the `main` branch with a malicious payload. 3. A user invokes the buyer workflow. 4. The Agent detects the operating system and runs the corresponding documented command. 5. The command downloads the current attacker-controlled r ...[truncated 918 chars]
Remediation
View remediation

T08 · Insecure Dependencies

Warning
Location
SKILL.md:41
Finding

External Skills Are Installed Without an Immutable Version

Content
View full analysis
Remediation
View remediation

T01 · Skill Instruction Hijacking

Error
Location
SKILL.md:57
Finding

Mutable Remote Documents Are Treated as Authoritative Implementation Instructions

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Chaining Abuse

High
Category
Tool Misuse
Confidence
99% confidence
Finding

The curl ... | sh pattern is a classic command-chaining anti-pattern because it executes network-fetched content immediately without giving the user a reliable inspection point. In this skill's context, the danger is elevated because the instructions are presented as the canonical setup path for payment infrastructure, making unsafe execution more likely and enabling full system compromise if the fetched content is malicious.

Content

Scanner excerpt · SKILL.md (reported line 33)May include surrounding context.

bash
# macOS / Linux
curl -sSL https://raw.githubusercontent.com/okx/onchainos-skills/main/install.sh | sh

# Windows PowerShell
irm https://raw.githubusercontent.com/okx/onchainos-skills/main/install.ps1 | iex

Rp1

Medium
Category
MCP Rug Pull
Confidence
91% confidence
Finding

The skill instructs users to run npx skills add okx/onchainos-skills without pinning a specific package version or immutable source. This can cause users to install a changed or compromised package at execution time, creating a supply-chain risk in a security-sensitive setup flow that manages payment tooling and potentially private-key-adjacent components.

Content

No source excerpt is available for this finding.

External Script Fetching

Low
Category
Supply Chain
Confidence
98% confidence
Finding

The skill includes remote script execution bootstrap commands that fetch installer code from GitHub at runtime. Even with a nearby note suggesting review, embedding these commands in an interactive skill materially increases the chance that users will execute unreviewed code, and if the upstream repository, branch, or delivery path is compromised, arbitrary code runs on the user's machine.

Content

Scanner excerpt · SKILL.md (reported line 33)May include surrounding context.

bash
# macOS / Linux
curl -sSL https://raw.githubusercontent.com/okx/onchainos-skills/main/install.sh | sh

# Windows PowerShell
irm https://raw.githubusercontent.com/okx/onchainos-skills/main/install.ps1 | iex

Static analysis

No suspicious patterns detected.