Back to skill

Security audit

PowPow Registro — Registre-se na Bolha, fixe seu diário de viagem no mapa, converse com humanos digitais

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed PowPow account-registration helper that handles sensitive signup and optional payment steps, but its behavior is coherent with that purpose.

Install only if you are comfortable letting the agent help create a real PowPow account and process your email, username, password, verification code, optional avatar, temporary login token, and optional payment order status. Payment credentials stay with Alipay or PayPal, but you can choose manual registration instead, and you should run cleanup after the assisted flow finishes.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (23)

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 19)May include surrounding context.

md
- Script principal `scripts/powpow-register.mjs`: **zero dependências**, requer **Node ≥ 18**. Antes de começar, rode uma vez `node scripts/powpow-register.mjs

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 65)May include surrounding context.

md
- Script principal `scripts/powpow-register.mjs`: **zero dependências**, requer **Node ≥ 18**. Antes de começar, rode uma vez `node scripts/powpow-register.mjs

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 76)May include surrounding context.

md
- Script principal `scripts/powpow-register.mjs`: **zero dependências**, requer **Node ≥ 18**. Antes de começar, rode uma vez `node scripts/powpow-register.mjs

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 84)May include surrounding context.

md
- Script principal `scripts/powpow-register.mjs`: **zero dependências**, requer **Node ≥ 18**. Antes de começar, rode uma vez `node scripts/powpow-register.mjs

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 100)May include surrounding context.

md
- Script principal `scripts/powpow-register.mjs`: **zero dependências**, requer **Node ≥ 18**. Antes de começar, rode uma vez `node scripts/powpow-register.mjs

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 111)May include surrounding context.

md
- Script principal `scripts/powpow-register.mjs`: **zero dependências**, requer **Node ≥ 18**. Antes de começar, rode uma vez `node scripts/powpow-register.mjs

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 122)May include surrounding context.

md
- Script principal `scripts/powpow-register.mjs`: **zero dependências**, requer **Node ≥ 18**. Antes de começar, rode uma vez `node scripts/powpow-register.mjs

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 126)May include surrounding context.

md
- Script principal `scripts/powpow-register.mjs`: **zero dependências**, requer **Node ≥ 18**. Antes de começar, rode uma vez `node scripts/powpow-register.mjs

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 133)May include surrounding context.

md
- Script principal `scripts/powpow-register.mjs`: **zero dependências**, requer **Node ≥ 18**. Antes de começar, rode uma vez `node scripts/powpow-register.mjs

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 134)May include surrounding context.

md
- Script principal `scripts/powpow-register.mjs`: **zero dependências**, requer **Node ≥ 18**. Antes de começar, rode uma vez `node scripts/powpow-register.mjs

Ae1

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 139)May include surrounding context.

md
- Script principal `scripts/powpow-register.mjs`: **zero dependências**, requer **Node ≥ 18**. Antes de começar, rode uma vez `node scripts/powpow-register.mjs

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The README promotes delegated registration, verification-code handling, and avatar upload by the agent, but does not clearly foreground that the agent will process personal account data including email, username, avatar, and one-time verification codes. In an account-creation workflow, this omission can mislead users about the sensitivity of the data being handled and reduce meaningful consent, increasing privacy and account-takeover risk if the skill or surrounding platform is misused or compromised.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
95% confidence
Finding

A skill that performs network actions and handles secrets via environment variables without declaring an explicit tool/permission scope creates an authorization gap: the agent may be allowed to use capabilities that are not transparently constrained or reviewable. In this skill, that matters because it collects registration data, transmits email addresses, handles passwords, and initiates payment-related flows, so missing scope declarations increase the risk of overbroad execution or unintended tool access.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill performs privacy- and system-impacting actions—submitting personal data, uploading an avatar, requesting email verification codes, creating payment orders, polling order status, and opening local/browser payment pages—without requiring an upfront, explicit user consent checkpoint before those actions begin. That can lead to users unintentionally authorizing external requests or payment flow initiation before they understand what data will be sent and what local actions will occur.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

SQP-3 applies to all file types and covers language/locale policy violations. The UX section prescribes a fixed service tone and response pattern in Portuguese, but does not offer any opt-in or alternative language behavior, which can improperly force a specific language on users.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The manifest sets a fixed language value of "pt", which indicates the skill is intended to operate in Portuguese only. There is no accompanying note offering a language choice or explaining a region-specific requirement, so this appears to be a locale-policy issue under the natural-language policy rules.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The file title and most instructions are in Portuguese, while user-facing strings and described UI text also appear in Chinese and English elsewhere in the document. This documents a multi-language interaction pattern without any opt-in, language selection, or justification for forcing those locales on users.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The manifest says the skill can handle optional subscription payment via Alipay/PayPal and explicitly frames unsupported areas narrowly, but this file additionally documents WeChat QR subscription and Huawei in-app purchase activation flows. Those capabilities are not justified by the stated purpose of this skill as managed channels and broaden the operational scope beyond the declared Alipay/PayPal handling.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The documentation explicitly states that the agent performs registration/login through public APIs and participates in handling authentication state via identity tokens and HttpOnly-cookie-based flows, but it does not require clear user consent, privacy notice, or confirmation before account-affecting actions. In an agent context, this is risky because the user may not understand that the assistant is initiating real account creation, login, token-bearing session transitions, and possibly payment-adjacent flows on their behalf.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

As linhas L03-L04 dizem para usar as capturas 'conforme a necessidade em tempo de execução' e enviá-las diretamente quando o usuário 'não faz ideia' do produto ou 'precisa ver' a interface. Isso não define gatilhos específicos nem limites claros, o que pode causar invocações excessivas ou inconsistentes em conversas comuns sobre o produto.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

Across comments, usage text, and runtime messages, the skill consistently forces Chinese-language interaction. The policy only allows locale constraints when users are given a choice or the restriction is clearly documented and justified, which is not present here.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
72% confidence
Finding

SQP-3 applies to natural-language policy issues in any file type, including markdown. The README presents the skill entirely in Portuguese and does not mention language choice or that the locale is intentionally limited, which can indicate a language/locale constraint without user opt-in.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
67% confidence
Finding

The file includes Chinese-only UI strings and a mainland China phone-number requirement, but does not explicitly state that the skill is intended only for that locale or provide a language choice. Under the policy, forcing a specific language or locale without opt-in can be a natural-language policy issue unless clearly justified as region-specific.

Content

No source excerpt is available for this finding.