ClawHub
Back to skill

Security audit

Shipyard email

Security checks for vulnerabilities and agentic risk

Overview

The skill provides purpose-aligned email setup with clear user controls, while relying on separate Shipyard scripts that were unavailable for review.

Review the full Shipyard plugin before installing; its executable scripts were outside this artifact. Keep keys out of chat and use only dummy content for diagnostic emails. Mail-tester receives test messages, and its [FAQ](https://mail-tester.com/faq) says results are accessible to anyone with the exact testing address.

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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Output HandlingUnvalidated Output Injection, Cross-Context Output, Unbounded Output
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Credential Access

High
Category
Privilege Escalation
Confidence
70% confidence
Finding

Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.

Content

Scanner excerpt · SKILL.md (reported line 15)May include surrounding context.

md
## Rules (always)

- You do the work. Never hand the student a command you could run.
- Never ask for keys or passwords in the chat and never read `~/.shipyard/secrets.env` or `.env*` files. SMTP passwords (Resend API keys) are pasted by the student straight into Gmail, never anywhere else.
- Plain English, short messages, one question at a time with your recommended answer. Explain any new word in five words or fewer ("MX record: where mail for the domain goes").
- Never delete or change an existing DNS record the student did not ask about (the website's records must keep working). Show what to add, not a whole new zone.
- Never suggest changing a sending setup that already works (for example records on `send.<domain>` made by another service). Report what you see; change it only if the student asks and something is actually broken.

The required mail-tester.com test lacks a privacy warning about sharing email contents and headers with a third-party diagnostic service.

Medium
Category
Output Handling
Confidence
91% confidence
Finding

The mandatory diagnostic test sends an email and its headers to a third party without explaining that disclosure or limiting the message to synthetic content. Deliverability testing is legitimate in this context, but users could expose private correspondence or application data unnecessarily; there is no evidence of malicious exfiltration.

Content

No source excerpt is available for this finding.

The required mail-tester.com tests lack a privacy warning about sending email content and metadata to a third-party service.

Medium
Category
Output Handling
Confidence
92% confidence
Finding

The reference requires sending messages to mail-tester.com and later extends testing to application-generated email, without a privacy warning or a requirement to use synthetic data. Although this is a legitimate diagnostic workflow, the service receives message bodies and headers, creating a potential disclosure if real application or personal content is used.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.