Robot

• Outcome: Pass
• Latest audit: Sep 10, 2026 · 8:00 PM UTC
• Version: 1.0.0

Overview

This robotics skill is coherent and not deceptive, but users should review its local note-taking and physical hardware safety assumptions before using motor-control examples.

Install only if you want a robotics assistant that may maintain local notes under ~/robot. Before applying generated motor, ESC, ROS, or industrial robot code to real hardware, test unloaded or in simulation, use a physical cutoff or emergency stop, and verify power/current limits and wiring yourself.

A.I.G

Vulnerability Patterns

• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally
• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (13)

Vague Triggers

Medium

• Confidence: 95% confidence
• Finding: The skill’s activation scope is very broad ('User needs robotics help') and covers code generation, hardware selection, and debugging without clear limits on when the skill should engage. In an agent setting, overly broad triggers can cause the skill to activate in unintended contexts and apply file-reading/writing behaviors or risky robotics guidance when the user did not explicitly request this specialized workflow.

Missing User Warnings

Medium

• Confidence: 93% confidence
• Finding: The skill instructs the agent to proactively read and write files under ~/robot/, including inventory, projects, corrections, and archives, without an explicit user-facing warning or consent step. This creates a local integrity and privacy risk because the agent may modify user files automatically, persist inaccurate or sensitive data, and do so in situations where the user did not expect filesystem side effects.

Ae4

Medium

• Category: analysis-evasion
• Confidence: 80% confidence
• Finding: Suspicious Unicode normalization or mixed-script content

Ae4

Medium

• Category: analysis-evasion
• Confidence: 80% confidence
• Finding: Suspicious Unicode normalization or mixed-script content

Ae4

Medium

• Category: analysis-evasion
• Confidence: 80% confidence
• Finding: Suspicious Unicode normalization or mixed-script content

Missing User Warnings

Medium

• Confidence: 91% confidence
• Finding: The BTS7960 example enables a 43A motor driver and provides immediate actuation logic without an explicit safety warning, interlock guidance, or safe startup state discussion. In a robotics skill, readers may directly wire and run the sample, which could cause sudden high-torque motion, pinched fingers, damaged hardware, or electrical hazards if a load is attached.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The ESC example sends arming and throttle-related commands but only notes that the arming sequence varies by manufacturer; it does not clearly warn that connected propellers or motors may spin unexpectedly during arming, calibration, or signal mismatch. In the robot/hardware context, this omission is dangerous because users often test immediately on a bench, where sudden rotation can cause laceration or projectile hazards.

Missing User Warnings

Medium

• Confidence: 94% confidence
• Finding: The line follower project includes battery power and motor-control guidance for physical hardware but omits basic safety warnings about LiPo handling, wiring mistakes, shorts, overcurrent, and unexpected actuator movement. In a robotics skill, users may directly replicate these instructions on real devices, so the absence of warnings increases the chance of injury or hardware damage even though the content is not overtly malicious.

Missing User Warnings

Medium

• Confidence: 97% confidence
• Finding: The self-balancing robot section provides live PID tuning and motor actuation code for an inherently unstable platform without warning that incorrect gains or bad calibration can cause sudden runaway motion, falls, or component damage. Because balancing robots can rapidly accelerate or topple during tuning, the missing safety guidance is more dangerous in this skill context than in a purely simulated or educational setting.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The ROS2 motor driver example subscribes to /cmd_vel and forwards commands directly to hardware over serial with no warning that running the node can cause immediate physical movement. In a robotics environment, command topics may be driven by other nodes, tests, or misconfigurations, so presenting this example without safety caveats can lead to unintended motion, collisions, or equipment damage.

Ae4

Medium

• Category: analysis-evasion
• Confidence: 80% confidence
• Finding: Suspicious Unicode normalization or mixed-script content

Missing User Warnings

Low

• Confidence: 97% confidence
• Finding: The skill instructs creation of directories and files in the user's home directory, which changes the local filesystem without any explicit warning, confirmation step, or note about where data will be stored. While the actions are limited and appear operational rather than malicious, silently causing persistent file changes can surprise users, overwrite expected conventions, and set a precedent for unsafe automation.

Missing User Warnings

Low

• Confidence: 94% confidence
• Finding: The archive rules direct moving projects and corrections into archive locations, which can rename or relocate user data without an explicit warning that files will be modified. In a robotics skill this is not inherently suspicious, but unannounced data movement can still cause confusion, broken references, or accidental loss of context if users did not intend the archival action.

Static analysis

No suspicious patterns detected.