Back to skill

Security audit

Vibe Coding 可行性评估

Security checks for vulnerabilities and agentic risk

Overview

The skill does what it claims, but its API-key handling can send an OpenAI key to DeepSeek by default and it does not clearly disclose the external data flow.

Review before installing. Use only a DeepSeek key unless the provider selection is fixed, do not run it in an environment containing unrelated OPENAI_API_KEY values, and avoid submitting confidential product plans, customer data, internal architecture, or secrets in the idea text until the skill clearly discloses and constrains its external API data flow.

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
scripts/evaluate_vibe.py:5
Finding

OpenAI API Key May Be Transmitted to the Default DeepSeek Endpoint

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Taint TrackingDirect Taint Flow, Variable-Mediated Taint Flow, Credential Exfiltration Chain
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Findings (7)

Tainted flow: 'req' from os.environ.get (line 16, credential/environment) → urllib.request.urlopen (network output)

Critical
Category
Data Flow
Confidence
90% confidence
Finding

Credentials or environment variables flow to a network sink. This is a high-confidence indicator of credential exfiltration.

Content

Scanner excerpt · scripts/evaluate_vibe.py (reported line 18)May include surrounding context.

python
], "temperature": 0.7}).encode()
    req = urllib.request.Request(f"{API_BASE}/chat/completions", data=payload,
        headers={"Content-Type": "application/json", "Authorization": f"Bearer {API_KEY}"})
    with urllib.request.urlopen(req, timeout=60) as r:
        return json.loads(r.read())["choices"][0]["message"]["content"]

def build_prompt(idea, skill_level):

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
92% confidence
Finding

The skill invokes a Python script and the static analysis indicates code capabilities including environment access and network use, but the manifest does not declare any tool scope such as permissions or allowed-tools. That creates an overbroad execution surface where the runtime may grant more capability than the skill documentation communicates, making unintended data access or outbound requests harder to review and contain.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The entire skill description and examples are presented only in Chinese, with no indication that users may interact in other languages or choose their preferred locale. Under the policy, language constraints should either be optional for the user or clearly documented as justified for a specific regional use case.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The phrase '描述你想做的功能或项目,我帮你评估' is a very general invitation that could overlap with many ordinary project-discussion requests. The document provides example prompts, but it does not clearly define specific trigger phrases, scope constraints, or negative examples to distinguish when this skill should activate versus when a general assistant should respond.

Content

No source excerpt is available for this finding.

Context-Inappropriate Capability

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

该技能会读取环境变量中的 API 凭证并向外部 LLM 服务发送请求,但描述中仅说明“快速评估可行性”,没有明确披露会进行外部网络调用和使用本地凭证。对使用者而言,这会造成能力边界不透明,可能在不知情情况下消耗第三方账户额度,或在受限环境中违反数据/网络使用预期。

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

用户输入的项目描述和技能水平会被直接发送到远程 LLM,但程序没有任何用户可见的提示、确认或敏感信息警告。项目描述中常包含未发布产品想法、内部架构、客户需求或密钥片段,这会导致无意的数据外传和合规风险;在该技能场景下,这种风险与其处理自由文本项目需求的上下文高度相关,因此更值得关注。

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
97% confidence
Finding

The system message explicitly instructs the model to answer in Chinese, which imposes a language choice regardless of user preference. The file does not offer a language option or indicate that this is a region-specific tool with a justified locale constraint.

Content

No source excerpt is available for this finding.