Multi-account Email Desk

• Outcome: Pass
• Latest audit: Oct 26, 2026 · 8:54 PM UTC
• Version: 1.0.1

Overview

The skill’s email access, sending controls, local storage, and optional reminders are disclosed and consistent with its purpose.

Connecting an account permits reading and sending mail and stores local copies. Protect the profile, token files, and mail cache; review recipients, attachments, and drafts before sending. Remove recurring jobs when no longer needed, and consider pinning the installer version.

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
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep
• Rogue Agent: Self-Modification, Session Persistence
• MCP Least Privilege: Underdeclared Capability, Wildcard Permission, Missing Permission Declaration

Findings (6)

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 21)May include surrounding context.

md

, then create a private profile outside the installed Skill directory. Credentials come from environment variables or an OAuth token file, never a chat transcript. Windows: use a user-private profile directory; Unix: mode 600. Do not place live profiles in GitHub or Skill ZIPs.

Choose IMAP for folders, flags and UID-based synchronization; POP3 supports download/UIDL polling and local search. SMTP sends prepared MIME drafts. SSL and STARTTLS verify certificates. OAuth2 XOAUTH2 supports existing access tokens and Google/Microsoft refresh-token profiles; an OAuth app registration and provider consent are needed first. App passwords use the password adapter. Provider policy may require enabling IMAP/SMTP.

First sync establishes a baseline and caches the most recent limit messages (default 50). Subsequent syncs fetch new messages in bounded batches; run again while remaining >0. search covers the local cache. For older mail, IMAP remote-search runs a server-side UTF-8 text/subject/from/to

MCP server referenced without pinned version: 'npx clawhub'.

Medium

• Category: MCP Rug Pull
• Confidence: 93% confidence

Finding

The documented command uses npx clawhub@latest, which resolves a mutable release and can execute newly published package code without a reproducible version pin. Although ClawHub is an installation CLI rather than an MCP server, the underlying supply-chain concern is valid: a compromised release could execute with the installing user's permissions and access workspace data or available mail credentials. The README shows no evidence of malicious intent or an actual compromise.

Content

No source excerpt is available for this finding.

MCP server referenced without pinned version: 'npx clawhub'.

Medium

• Category: MCP Rug Pull
• Confidence: 96% confidence

Finding

The installation command uses npx clawhub@latest, which selects a mutable release rather than a reviewed, fixed version and may download and execute it with the user's permissions. A compromised future release could access workspace files or locally accessible email credentials, increasing the consequences in this email-management context. The finding incorrectly describes ClawHub as an MCP server, but the underlying unpinned executable dependency concern is valid; there is no evidence here of an actual compromise.

Content

No source excerpt is available for this finding.

Skill declares no tool scope ('permissions' or 'allowed-tools') but code capabilities were detected: env, file_read, file_write, network, shell.

Medium

• Category: MCP Least Privilege
• Confidence: 70% confidence

Finding

Without declared permissions the skill's intent is opaque and cannot be validated.

Content

No source excerpt is available for this finding.

Session Persistence

Medium

• Category: Rogue Agent
• Confidence: 80% confidence

Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · references/notifications.md (reported line 3)May include surrounding context.

md

# Host scheduling and delivery

These scripts keep state and queues; Python by itself does not create a timer. For a user-requested recurring reminder, use the installed OpenClaw scheduler, inspect its current help, and create a job for THIS agent/workspace and the requested chat/account only. Get the timezone, frequency and recipient from the user's request; ask only for missing details. Verify the returned job ID and next run. Never reuse another agent's mailbox credentials or destination.

The job runs poll (email: sync) with the same private profile and workspace. If there are no pending notices, finish quietly. Deliver pending notices through the host's configured messaging tool. After each successful delivery, call ack with only its noticeId. If delivery fails, leave it pending. Maintain the noticeId in the host delivery ledger to avoid duplicate delivery after a retry; this is at-least-once delivery, not a cross-service transaction. Never acknowledge before sending. Manual tests

Scope Creep

Low

• Category: Excessive Agency
• Confidence: 70% confidence

Finding

Skill's behavior or capabilities extend beyond its stated purpose. Scope creep allows an agent to perform actions unrelated to its documented functionality, increasing the attack surface.

Content

Scanner excerpt · LICENSE (reported line 7)May include surrounding context.

text

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.