Back to skill

Security audit

Openclaw Performance Skill

Security checks for vulnerabilities and agentic risk

Overview

This skill is not clearly malicious, but it needs Review because its main scripts are incomplete and its optimizer can change local OpenClaw settings without enough user control.

Review before installing. Do not run the one-click optimizer unless you are comfortable with it changing your OpenClaw config; make your own backup first. Treat the monitoring numbers as demo-quality unless replaced with real OpenClaw telemetry, and be aware that several advertised scripts appear incomplete and may not run.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (18)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 43)May include surrounding context.

md
node scripts/diagnose.js

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 46)May include surrounding context.

md
node scripts/optimize.js

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 49)May include surrounding context.

md
node scripts/monitor.js

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill advertises '一键优化' and '监控部署' as automatic actions, but the documentation does not clearly disclose what configuration, prompts, system settings, or monitoring components will be changed. In an agent skill context, vague automation that can modify local application or system state increases the risk of unintended configuration drift, service instability, or privacy-impacting monitoring changes, especially if users run the referenced scripts with trust.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

This markdown file contains user-facing instructions exclusively in Chinese from title to conclusion, with no indication that the skill or documentation is intentionally limited to Chinese-speaking users. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This JavaScript file presents its title, console output, and embedded prompt/instructions entirely in Chinese. Because the skill does not offer any language selection or indicate that Chinese is a required regional context, it effectively forces a specific language without user opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The file’s description and all console-facing strings are presented in Chinese, which imposes a specific language on users. There is no opt-in, locale selection, or documented justification that this script is intended only for a Chinese-speaking or region-specific audience.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

This code presents its title, status messages, diagnostics, and recommendations entirely in Chinese, starting with the header comment and continuing throughout the script. Under the policy rule for natural-language violations, forcing a specific language without user opt-in or a documented regional justification is a locale-policy issue.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The file's natural-language strings and comments present the tool entirely in Chinese, including the user-facing console interface, with no indication that language selection is optional. Under the policy, forcing a specific language without user opt-in is a locale/language policy violation unless the constraint is clearly documented and justified.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The script presents itself as a performance monitor for OpenClaw, but the request count and response-time data are explicitly simulated with Math.random() rather than collected from the target system. This can mislead operators into trusting fabricated telemetry, causing bad operational decisions, missed incidents, or false confidence about system health.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

User-facing messages, comments, and generated script text are written exclusively in Chinese throughout the skill, with no indication that language is configurable or based on user preference. This can violate language/locale policy when a skill implicitly forces one language without opt-in.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The script writes directly to the user's ~/.openclaw/config.json after merging settings, without confirmation, dry-run mode, or explicit consent. In a skill context, silently changing user configuration can disrupt service behavior, weaken user-chosen safeguards, or create persistence that the user did not intend.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The method heading and log output describe this phase as validating whether the optimization worked. However, the '系统响应测试' check merely measures a 100ms setTimeout delay in the Node process, which does not test OpenClaw responsiveness or optimization impact. This is an active mismatch between the documented intent of the verification step and the code's actual behavior.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

Line L15 states '优先使用Claude Haiku模型以获得更快响应', which imposes a specific model preference as a natural-language instruction. This can be a language/locale-style policy issue because it fixes a user-facing configuration choice without documenting opt-in or alternatives.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

This JSON report uses Chinese natural-language strings throughout, including the rating, result names/descriptions, and recommendations, with no indication that the language was user-selected or that the report is intended only for a Chinese locale. That can violate a language/locale policy when skills are expected to avoid forcing a specific language without user opt-in.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
89% confidence
Finding

The module comment describes the file as an OpenClaw performance diagnosis script, which implies observation and reporting behavior. Later, the script creates a reports directory and writes a JSON report to disk at L485-L491, adding a persistent side effect not reflected in the top-level documentation.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
85% confidence
Finding

The script creates executable files in the current workspace without prior warning or consent. While the generated files are not immediately executed, dropping runnable scripts into a user workspace can surprise users, clutter trusted project directories, and increase the chance of accidental execution later.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The parent script generates a health-check script that executes subprocess commands like 'openclaw --version' and 'openclaw doctor', but does not clearly warn the user that runnable command-executing code is being created. This is dangerous in a skill setting because users may later run the generated script assuming it is harmless, leading to execution of external commands in their environment.

Content

No source excerpt is available for this finding.