Back to skill

Security audit

Skill Plant Care

Security checks for vulnerabilities and agentic risk

Overview

The skill is a coherent plant-care guide with no executable code, but it gives specific pesticide and fungicide advice without enough safety warnings for users.

Review the chemical-treatment sections carefully before installing. The skill appears non-executable and purpose-aligned, but users should be told to follow local product labels, use appropriate protective equipment, keep chemicals away from children and pets, and prefer lower-risk controls when possible.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (5)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The documented triggers are broad enough to match common plant-related requests such as photos, care questions, and diagnosis queries, which can cause the skill to activate when the user did not explicitly intend to invoke it. This is not a code-execution issue, but it can lead to unintended skill routing, confusing responses, or overshadowing other more appropriate skills.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The guide recommends multiple pesticides and fungicides such as imidacloprid, abamectin, malathion, carbendazim, chlorothalonil, sulfur powder, and acetamiprid without any safety warnings, PPE guidance, dilution instructions, ventilation advice, storage/disposal cautions, or instruction to follow local product labels. In a consumer-facing plant-care troubleshooting document, this omission can lead users to misuse toxic chemicals around people, pets, food areas, or pollinators, causing poisoning or environmental harm.

Content

No source excerpt is available for this finding.

Vague Triggers

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The example phrase 'Help me identify this plant' is ordinary conversational language and may be reused by many benign interactions, making it a weak trigger anchor if used operationally. Generic phrases increase the chance of accidental invocation and reduce separation between this skill and adjacent capabilities.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
90% confidence
Finding

This markdown file provides care instructions for many plants, including some explicitly noted as hazardous, but it does not include a user-facing warning about pet/child toxicity or skin irritation risks. For example, Rubber Plant is described as having 'white milky sap,' which can irritate skin, yet no caution accompanies the care guidance.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
98% confidence
Finding

At the Giant Taro entry, the file explicitly notes the plant is 'somewhat toxic' and then immediately gives care instructions without any warning about ingestion or contact hazards. Under the markdown-file warning criteria, behaviour affecting user safety should be disclosed when the skill encourages plant care actions.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.