Back to skill

Security audit

Ai Intelligent Knowledge Qa

Security checks for vulnerabilities and agentic risk

Overview

The skill is not clearly malicious, but it asks users to run unpinned external Python code that is not included in the reviewed package.

Install only after reviewing the exact GitHub revision and dependency set you will run. Prefer a pinned commit or signed release, run it in an isolated environment, and avoid exposing sensitive files, API keys, or production credentials until the external source has been vetted.

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • 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
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Findings (1)

T03 · Remote Payload Retrieval and Execution

Error
Location
SKILL.md:39
Finding

Unpinned Remote Code and Dependency Retrieval Followed by Execution

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 39-42
Vulnerability Type: Unpinned remote payload retrieval and insecure dependency installation
Risk Level: High

Complete Code Snippet:

bash
git clone https://github.com/openclaw-skills/ai-intelligent-knowledge-qa
cd ai-intelligent-knowledge-qa
pip install -r requirements.txt
python app.py

Technical Analysis

The audited package contains only documentation and metadata; it does not include the referenced app.py, requirements.txt, or the advertised application implementation. The installation instructions clone the current state of an external repository without pinning an immutable commit, release, or cryptographic digest. They then install dependencies from an externally retrieved requirements file and execute the downloaded Python application.

This creates a time-of-check/time-of-use supply-chain risk: the code reviewed in this package does not determine the code that users will ultimately run. The external repository's default branch can change after this skill is reviewed. Furthermore, the dependency installation command does not show the use of a lockfile, package hashes, trusted package index restrictions, or integrity verification.

No evidence establishes that the current remote repository or its dependencies are malicious. The vulnerability is the unsafe, mutable retrieval and execution workflow itself.

Attack Path

  1. An attacker compromises the referenced repository, an authorized maintainer account, or an upstream dependency selected by requirements.txt.
  2. The attacker adds malicious code to the repository's default branch or causes dependency resolution to select a malicious package version.
  3. A user follows the documented installation procedure after the remote content has changed.
  4. git clone retrieves the attacker-controlled content without checking an approved commit hash or signature.
  5. `pip install - ...[truncated 1052 chars]
Remediation
View remediation

Remediation Suggestions

  1. Include the complete application source and dependency manifests in the reviewed skill package so its effective runtime behavior can be audited.
  2. If external retrieval is unavoidable, pin the repository to a reviewed immutable commit hash rather than cloning the mutable default branch.
  3. Verify the repository revision using a trusted signature or an independently distributed cryptographic digest before installation or execution.
  4. Use a fully resolved dependency lockfile with exact versions and cryptographic hashes. Install with hash enforcement, such as pip install --require-hashes, where supported.
  5. Restrict dependency resolution to explicitly approved package indexes and review all direct and transitive dependencies.
  6. Separate retrieval, verification, installation, and execution into distinct steps. Abort if any integrity or signature check fails.
  7. Execute the application in an isolated, least-privilege environment with minimal filesystem access, no unnecessary credentials, and restricted outbound network connectivity.
  8. Add automated source, dependency, and artifact scanning to the release process, and publish reproducible, versioned releases instead of relying on a mutable branch.
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 (2)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The natural-language metadata describes the skill in Chinese and explicitly tags it as "chinese," which signals a language-specific experience. There is no accompanying note that the skill supports user language choice or that the Chinese-only scope is intentional and justified for a region-specific use case.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

The natural-language content of the skill, including its description and usage sections, is presented only in Chinese. Under the policy criteria, forcing a specific language without user opt-in or justification can be a locale-policy violation.

Content

No source excerpt is available for this finding.