Back to skill

Security audit

EO Quickstart

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed Chinese quickstart guide for installing and trying the EO collaboration plugin, with no hidden execution or destructive behavior in the reviewed artifact.

Before installing, confirm that eo-collaboration is the plugin you intend to use and that it comes from the expected ClawHub/OpenClaw publisher or registry entry. Treat the external plugin as a separate trust decision because this skill only documents how to install and use it.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
Findings (1)

T08 · Insecure Dependencies

Warning
Location
SKILL.md:27
Finding

Unpinned Third-Party Plugin Installation

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 27–28 and 74–77
Vulnerability Type: Unverified and unpinned third-party dependency installation
Risk Level: Medium

Vulnerable Code

bash
openclaw plugins install eo-collaboration
# or
clawhub install eo-collaboration

The installation instruction is repeated later:

bash
openclaw plugins install eo-collaboration

Technical Analysis

The Skill directs users to install the externally maintained eo-collaboration plugin by package name without specifying an immutable version, content digest, checksum, signature-verification procedure, or verified publisher identity.

Consequently, the installed artifact is resolved remotely at installation time and may differ from the artifact available when this Skill was audited. The project contains only SKILL.md; it does not include the plugin implementation, so the external component's behavior cannot be validated from this repository.

This creates a supply-chain trust boundary. Compromise of the package registry, publisher account, dependency name, or a mutable plugin release could cause users to install code that does not match the documented functionality.

Attack Path

  1. An attacker compromises the plugin publisher account or relevant distribution channel, or otherwise causes a malicious artifact to resolve under eo-collaboration.
  2. The attacker publishes or substitutes a modified plugin release.
  3. A user follows the instructions in SKILL.md and runs one of the unpinned installation commands.
  4. The package manager retrieves the current remote artifact without verification against a documented immutable digest.
  5. The malicious plugin executes when loaded or invoked in the OpenClaw Agent environment.

Impact Assessment

Successful exploitation could allow malicious plugin code to operate with the permissions granted to the OpenClaw Agent or plugin runtime. Depending on those runtime permissions, potential effect ...[truncated 326 chars]

Remediation
View remediation

Remediation Suggestions

  1. Pin the plugin to a reviewed, immutable version rather than installing the latest package by name.
  2. Prefer an immutable artifact digest or cryptographic hash when the package manager supports it.
  3. Require signature verification and document the expected signer or verified publisher identity.
  4. Link only to an authenticated official registry and avoid alternate installation channels unless both provide equivalent provenance guarantees.
  5. Document the plugin's requested permissions and instruct users to grant only the minimum capabilities required.
  6. Record the reviewed plugin version, checksum, publisher, and review date in SKILL.md.
  7. Consider vendoring or separately auditing the exact plugin artifact when stronger assurance is required.
  8. Apply the same controls to both occurrences of the installation instruction, including the repeated command at lines 74–77.
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

This markdown skill presents all user-facing guidance in Chinese, including the title, instructions, and examples, with no indication that the user can choose another language. That creates a natural-language policy concern because it effectively forces a specific language without user opt-in.

Content

No source excerpt is available for this finding.