Security audit

Physical Mail Action Sorter

Security checks for vulnerabilities and agentic risk

Overview

This is a prompt-only mail-organizing skill that asks users to avoid sharing sensitive identifiers and does not include code, network access, credentials, or persistence.

Before installing, note that this skill may discuss bills, notices, healthcare, insurance, or government mail at a high level. Use only sender names and general envelope clues, and do not share account numbers, IDs, payment details, passwords, one-time codes, or private case numbers.

A.I.G

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

SkillSpector

Vulnerability Patterns

Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
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

Findings (5)

Natural-Language Policy Violations

Medium

Confidence: 96% confidence
Finding: The acceptance criteria explicitly require 'Public-facing documentation is English only.' This is a natural-language locale restriction, and the file does not offer users a language choice or provide a clear region-specific justification for the constraint.

Natural-Language Policy Violations

Medium

Confidence: 95% confidence
Finding: The clean scan evidence states 'No CJK characters; English/ASCII only.' This enforces a specific language/character set policy without indicating user consent or a necessary compliance reason, which falls under the language/locale policy violation category.

Vague Triggers

Medium

Confidence: 89% confidence
Finding: The trigger phrase "sort my mail" is short and broadly phrased, making it likely to overlap with ordinary user speech rather than a clearly delimited skill invocation. The manifest does not provide negative examples or tighter activation constraints to distinguish when this skill should activate versus when it should not.

Vague Triggers

Medium

Confidence: 82% confidence
Finding: Phrases like "physical mail pile," "paperwork pile," and "organize envelopes" describe a topic but do not clearly define invocation boundaries. Without explicit constraints or negative examples, these keywords could cause unintended activation for general organization or paperwork discussions.

Natural-Language Policy Violations

Low

Confidence: 91% confidence
Finding: The manifest sets the skill language to "en," which can indicate an English-only constraint. There is no accompanying note that this is optional, user-selectable, or justified by a region-specific requirement.

Static analysis

No suspicious patterns detected.