Back to skill

Security audit

Bingo Email

Security checks for vulnerabilities and agentic risk

Overview

The skill is a real email-management tool, but it needs review because it can access, send, forward, and permanently delete email with weak install and transport safeguards.

Install only if you are comfortable giving this skill access to your mailbox and stored app password or authorization code. Avoid the standalone curl install path, prefer the reviewed packaged script, keep the credential file private, do not use encryption = none, and manually review recipients, message bodies, forwards, reply-all actions, and deletes before letting an agent execute them.

Vulnerability Patterns
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Insecure DependenciesIntroduces malicious components through unsafe dependency sources
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • 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
Findings (3)

T03 · Remote Payload Retrieval and Execution

Error
Location
README.md:45
Finding

Mutable Remote Script Is Downloaded and Executed Without Integrity Verification

Content
View full analysis
Remediation
View remediation
/scripts/bingo_email.py ``` 4. Publish a SHA-256 checksum through a separately protected release process and verify it before execution: ```bash echo " bingo_email.py" | sha256sum --check - ``` 5. Prefer signed release artifacts and verify the maintainer's signature against a documented trusted key. 6. Use `curl -fL` so HTTP errors cause installation to fail instead of potentially saving an error response as the target file. 7. Do not make or execute the file until all integrity checks have succeeded. 8. Clearly document that initialization handles mailbox credentials and must only be performed with a verified build. ]]>

T08 · Insecure Dependencies

Warning
Location
SKILL.md:21
Finding

Security-Critical Dependency Is Installed Without Version or Hash Pinning

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
scripts/bingo_email.py:205
Finding

Configuration Allows Mailbox Credentials and Email Data to Be Sent Over Plaintext Connections

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
Findings (15)

Missing User Warnings

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The delete command immediately marks the message as \Deleted and expunges it, resulting in permanent removal without any confirmation, undo path, or safer trash-first behavior. In an agent-driven workflow, a mistaken instruction, prompt ambiguity, or malicious coercion could irreversibly destroy user data.

Content

No source excerpt is available for this finding.

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
80% confidence
Finding

Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

Content

Scanner excerpt · README.md (reported line 17)May include surrounding context.

md
- 🔧 **126/163 兼容** — 自动处理 IMAP ID (RFC 2971) 要求,告别 "Unsafe Login"
- 🇨🇳 **中文零乱码** — 草稿箱使用 base64 编码(非 quoted-printable)
- 📦 **轻量依赖** — 仅需 `imapclient`,纯 Python 无编译
- 🔒 **安全设计** — 配置文件自动 chmod 600、密码输入无回显、.gitignore 防泄露
- 📩 **完整功能** — 读取、发送、回复、转发、草稿、删除、移动
- 🧵 **线程归并** — 自动设置 In-Reply-To / References 头
- 🤖 **AI 友好** — `check` 命令一键诊断,Agent 无需人工干预

Sudo/Root Execution

Medium
Category
Privilege Escalation
Confidence
80% confidence
Finding

Commands invoke sudo or root privileges. Verify this elevated access is necessary and justified.

Content

Scanner excerpt · SKILL.md (reported line 41)May include surrounding context.

md
- 🔧 **126/163 兼容** — 自动处理 IMAP ID (RFC 2971) 要求,告别 "Unsafe Login"
- 🇨🇳 **中文零乱码** — 草稿箱使用 base64 编码(非 quoted-printable)
- 📦 **轻量依赖** — 仅需 `imapclient`,纯 Python 无编译
- 🔒 **安全设计** — 配置文件自动 chmod 600、密码输入无回显、.gitignore 防泄露
- 📩 **完整功能** — 读取、发送、回复、转发、草稿、删除、移动
- 🧵 **线程归并** — 自动设置 In-Reply-To / References 头
- 🤖 **AI 友好** — `check` 命令一键诊断,Agent 无需人工干预

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
95% confidence
Finding

The skill clearly enables shell execution, filesystem access, and live IMAP/SMTP network operations, yet it declares no explicit tool scope or permission boundaries. In an agent environment, that omission increases the chance the skill is invoked with broader-than-necessary authority, allowing sensitive mailbox access, config reads, message sends, and file writes without clear sandbox expectations.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger text is extremely broad, covering nearly any mention of email-related words, which can cause the agent to select this skill in casual discussion or ambiguous contexts. Because the skill can read mail, send messages, and manipulate folders, over-triggering creates a real risk of unintended privacy-impacting or externally visible actions.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The documented automatic flow says the agent should directly proceed to the requested mail operation after a successful check, but it does not require a privacy warning or confirmation before accessing mailbox contents. In a real deployment, this can lead to silent retrieval of sensitive email data when the user did not realize the skill would immediately connect and read their account.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The send/reply/forward commands are documented as direct actions, but the skill text does not impose a clear warning or confirmation requirement before outbound messages are transmitted. Since email sending is externally visible and may expose recipients, quoted thread history, and attachments/content, accidental or automated execution can cause data leakage or reputational harm.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The reply-all flow automatically expands recipients from the original To/Cc headers and sends without any explicit confirmation or preview of who will be copied. In an agent-assisted context, this increases the risk of unintended disclosure of sensitive information to extra recipients, especially when the user asked only for a reply in natural language.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

Forwarding includes the original message body and sends it immediately without a warning or confirmation about the content being transmitted. In an agent context, that can leak confidential email content, credentials, internal discussions, or personal data to an unintended third party if the forwarding instruction is ambiguous or manipulated.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill description frames the tool as mailbox reading/sending/management but does not clearly disclose that it can permanently delete and move messages. In an agent setting, incomplete capability disclosure can cause unsafe invocation by a user or orchestrator, leading to unintended destructive mailbox actions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
76% confidence
Finding

The skill documentation and examples are entirely presented in Chinese and emphasize Chinese-specific behavior, but there is no indication that the user can choose another language or that the locale restriction is intentional and documented as such. This can violate language-choice expectations unless the skill is explicitly scoped to Chinese-speaking users.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
91% confidence
Finding

The module description and all user-facing CLI messages are written in Chinese, and the tool does not indicate any language choice or opt-in. This creates a natural-language locale constraint that may violate policy where skills should not force a specific language without user selection or documented justification.

Content

No source excerpt is available for this finding.

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

Scanner excerpt · scripts/bingo_email.py (reported line 591)May include surrounding context.

python
def addr_display(envelope, field):
    """安全地获取地址显示名"""
    addrs = getattr(envelope, field) or []
    if not addrs:
        return ""
    first = addrs[0] if isinstance(addrs, (list, tuple)) else addrs

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

Scanner excerpt · scripts/bingo_email.py (reported line 608)May include surrounding context.

python
def addr_display(envelope, field):
    """安全地获取地址显示名"""
    addrs = getattr(envelope, field) or []
    if not addrs:
        return ""
    first = addrs[0] if isinstance(addrs, (list, tuple)) else addrs

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

cmd_check is documented as a diagnostic for environment readiness, but its structured TOML path requires account.password via has_pwd = bool(acct.get("password")) even though the configuration schema stores passwords under imap.password and smtp.password. This makes the check's stated intent contradict its actual behavior by falsely reporting incomplete configuration for valid configs.

Content

No source excerpt is available for this finding.