Back to skill

Security audit

Vision-Action-Evolution Loop

Security checks for vulnerabilities and agentic risk

Overview

The skill is not obviously malicious, but it describes robot-arm execution and self-updating behavior without clear safety gates or user-controlled limits.

Review carefully before installing for any environment connected to real hardware. Use only in simulation or a supervised, physically safe workcell until pre-execution human approval, collision checks, force and speed limits, emergency stop, confidence thresholds, update logging, and rollback controls are clearly defined. Pin and review the related skills before installing them.

Vulnerability Patterns
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • 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
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Findings (2)

T08 · Insecure Dependencies

Warning
Location
README.md:41
Finding

Unpinned Third-Party Skill Installation Creates a Supply-Chain Risk

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:39
Finding

Physical Robot Actions Are Defined Before Human Safety Approval

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (8)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The file introduces '心跳触发任务' and '每次心跳' as activation conditions, but does not define what constitutes a heartbeat, how often it occurs, or under what system context it runs. This ambiguity can cause unintended or overly frequent invocation because the trigger scope and exclusions are not specified.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The operational instructions in this file are written entirely in Chinese, which effectively forces a specific language for users or operators without any opt-in or documented locale requirement. Under the policy, language constraints should either provide user choice or be clearly justified as region-specific.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The README explicitly promotes a loop that progresses from perception to physical execution and self-evolution, but it provides no safety boundaries, human-approval requirements, or warnings about real-world actions. In an agent-skill ecosystem, this omission can encourage unsafe deployment of autonomous behaviors that affect physical systems or external environments without adequate oversight.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill includes physical robot actions such as grasping, folding, and assembly without any operator safety notice, environment constraints, or requirement for simulation-first validation. Because the skill is framed as an embodied execution loop, users could apply it to real hardware and cause equipment damage or bodily harm if planning or perception is wrong.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill explicitly describes an iterative loop that updates its world model and parameters over time, but it provides no warning, guardrails, approval checkpoints, or rollback guidance for that self-evolving behavior. In a robotics/planning context, silent self-modification can cause drift in task behavior, unsafe planning decisions, or propagation of bad feedback into future runs.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The verification protocol headings and checklist content are written in Chinese, and the file provides no indication that language selection is optional or that the skill is intentionally limited to a Chinese-speaking context. Under the stated 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

Low
Category
Not specified by scanner
Confidence
95% confidence
Finding

The natural-language content of the skill description is presented only in Chinese, and there is no indication that users may choose another language or that the skill is intentionally limited to a Chinese-speaking context. Under the stated policy, forcing a specific language without opt-in is a natural-language policy concern.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

Most of the natural-language content, including the description, headings, and usage guidance, is written in Chinese, but the file does not state that Chinese is required or offer an alternative language option. This can violate language/locale policy when a skill implicitly forces a specific language without user opt-in.

Content

No source excerpt is available for this finding.