Back to skill

Security audit

eBay Batch Registration

Security checks for vulnerabilities and agentic risk

Overview

This skill is a review concern because it guides bulk eBay/Payoneer account registration while storing sensitive identity, credential, banking, and card data in plaintext and giving anti-detection operational advice.

Install only after a careful review. This skill should not be used for platform-policy evasion or unauthorized bulk account creation. At minimum, remove proxy/risk-control evasion guidance, remove CVV storage, move credentials to a secrets manager, encrypt or avoid local bulk identity records, mask reports by default, define retention and deletion rules, and make scheduling/notifications explicit and opt-in.

Vulnerability Patterns
  • 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
  • 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
Findings (1)

T09 · Insecure Skill Coding Practices

Error
Location
SKILL.md:154
Finding

Plaintext Storage of Identity, Authentication, Banking, and Payment-Card Data

Content
View full analysis

Vulnerability Details

File Location: SKILL.md, lines 154-187
Vulnerability Type: Plaintext storage and aggregation of highly sensitive data
Risk Level: High

Vulnerable Code Snippet

text
==== Batch Registration Data Template ====
Legal Person ID: {id}
Legal Person Name: {name}
Legal Person Name in Pinyin:
Date of Birth:
Identity Card Number:

Email: {auto_generated}
Email Password: {auto_generated}

eBay Account: {same as email}
eBay Password: {auto_generated}

eBay Username: {randomly generated by script}
Payoneer Email: {same as above}
Payoneer Password: {auto_generated}

eBay Phone Number: {record after receiving SMS}
Payoneer Phone Number: {record after receiving SMS}

eBay Credit Card Number:
eBay Credit Card Expiration Date:
eBay Credit Card CVV:

Payoneer Bank Name:
Payoneer Bank Account Number:
Payoneer Bank Address:

Company Name in Chinese:
Company Name in English:
Unified Social Credit Code:
Company Registration Date:
Company Postal Code:
Company Address in Chinese:
Company Address in English:

The snippet above is an English rendering of the plaintext template in SKILL.md; the original fields appear at the cited location.

Technical Analysis

The Skill instructs operators to place government identity details, account identifiers, three passwords, telephone numbers, full payment-card details—including CVV—and bank-account information into a plaintext registration file. The documented directory layout creates one such file for each of up to 50 legal persons under a predictable desktop directory.

This design violates data-minimization and secure-secret-storage principles. It aggregates several independent authentication and financial factors in one location, substantially increasing the value of a single file-system compromise. The Skill specifies no encryption at rest, credential-vault integration, field-level access controls, masking, retention period, secure deletion, backup exclusions, or file-permission requirem ...[truncated 1959 chars]

Remediation
View remediation

Remediation Suggestions

  1. Do not store CVV values. Remove the CVV field from all templates and ensure it is used only transiently during an authorized payment interaction.
  2. Move credentials into a secret store. Store passwords in an operating-system credential vault or managed secrets service rather than text, JSON, spreadsheet, or report files.
  3. Encrypt required records at rest. Use authenticated encryption with keys kept separately from the data. Apply field-level encryption to identity, card, and bank-account data.
  4. Separate data domains. Do not aggregate identity records, passwords, recovery channels, and payment details in the same file or datastore.
  5. Apply least-privilege access controls. Restrict the data directory to the designated operator account and deny access to unrelated users and services.
  6. Prevent secondary copies. Exclude sensitive records from cloud synchronization, plaintext backups, temporary files, logs, generated reports, and source-control repositories.
  7. Mask output fields. Queue files, dashboards, logs, and Excel reports should display only redacted values, such as the last four digits of an account number.
  8. Minimize collection and retention. Collect only fields required for the immediate authorized operation, define short retention periods, and securely delete records once no longer needed.
  9. Add operational safeguards. Document file-permission requirements, incident-response procedures, access auditing, and secure disposal requirements.
  10. Protect generated reports. Ensure report exports omit passwords, CVVs, full card numbers, identity numbers, and complete bank-account details.
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (5)

Missing User Warnings

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill defines bulk storage and processing of highly sensitive personal, identity, financial, and account data, including ID numbers, business registration details, bank data, payment card fields, phone numbers, and account credentials, but provides no privacy, encryption, access-control, retention, or redaction safeguards. At 20–50 person scale, a compromise or accidental exposure would create a severe concentration of identity and financial risk.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

High
Category
Not specified by scanner
Confidence
99% confidence
Finding

The skill advises using different proxy IPs for each legal person specifically to avoid registration blocks and platform risk controls. In a bulk-registration tool, proxy rotation for this purpose is a classic anti-detection technique that enables large-scale account creation while concealing common origin, making abuse harder to detect and investigate.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill explicitly introduces randomized 2–4 hour delays to 'simulate natural rhythm' and reduce platform risk detection during batch account registration. In context, this is not neutral job scheduling: it is operational guidance for evading eBay anti-abuse controls while automating account creation at scale, which materially increases the likelihood of stealthy fraud or policy-violating activity.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The report-generation workflow exports identity-linked eBay and Payoneer account information, status, timelines, and failure data into Excel files without clear warning or protective controls. Spreadsheet exports are easy to copy, sync, email, or open on unmanaged endpoints, substantially increasing the risk of bulk data leakage.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Low
Category
Not specified by scanner
Confidence
79% confidence
Finding

The skill adds automatic next-day resumption and external failure notifications even though these operational capabilities are not clearly disclosed in the manifest. In this context, hidden persistence and out-of-band signaling can make large-scale automated registration more durable and less observable to reviewers, increasing abuse potential.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.