Back to skill

Security audit

SKILL2

Security checks for vulnerabilities and agentic risk

Overview

This is a local password and token generator with no evidence of network access, persistence, file writes, or hidden behavior, but users should avoid its unsafe password-entry examples and treat strength ratings as rough guidance.

Install only if you are comfortable with a local CLI that prints generated secrets to the terminal. Do not enter real passwords as command-line arguments or via literal echo pipelines; use safer hidden input methods where possible. Treat the strength checker as a rough composition estimate, not a reliable measure of whether a human-created password is hard to guess.

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 (2)

T09 · Insecure Skill Coding Practices

Warning
Location
passgen.py:147
Finding

Misleading Password-Strength Estimation Can Overrate Predictable Passwords

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Note
Location
SKILL.md:92
Finding

Documented Password Input Method Can Expose Credentials Through Shell History

Content
View full analysis
/passgen.py strength ``` `README.md`: ```bash echo "P@ssw0rd" | python3 passgen.py strength # 从管道读,避免进历史 ``` ### Technical Analysis The documentation states that piping a password through `echo` avoids placing it in command history. This is incorrect because the password remains literal text in the shell command itself. Interactive shells may record the entire `echo "password" | ...` command in their history files. Depending on the operating system and shell, literal command-line secrets may also be exposed temporarily through process inspection, terminal logging, session recording, monitoring software, or diagnostic collection. Although `passgen.py` only reads stdin and does not itself persist the submitted password, the recommended invocation method can cause exposure before the password reaches the program. ### Attack Path 1. A user replaces the example value with a real password and runs the documented pipeline. 2. The shell records the complete command, including the plaintext password, in its history. 3. A local user, administrative process, backup system, support bundle, or later compromise obtains access to the shell-history data. 4. The plaintext password is recovered. 5. If the password is active or reused, the attacker authenticates to the associated account or service. ### Impact Assessment This issue does not directly elevate privileges in the Skill process. It can disclose a plaintext credential to principals or systems that can inspect shell history, command-line activity, terminal logs, or backups. The resulting privilege scope depends on the exposed credential. An attacker may obtain access to any account or service protected by that ...[truncated 55 chars]
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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill documentation is written entirely in Chinese, and there is no indication that users may choose another language or that the skill is intentionally limited to a Chinese-speaking audience for a documented reason. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation.

Content

No source excerpt is available for this finding.