Back to skill

Security audit

Free AI Bot

Security checks for vulnerabilities and agentic risk

Overview

This skill is a Review item because its default routing can send user prompts to Cloudflare or Groq after local failure without a clear per-request cloud opt-in.

Install only if you are comfortable with prompts being sent to Cloudflare or Groq when cloud credentials are configured. Use --provider ollama for local-only use, avoid submitting secrets or private data, and note that the included Python script appears broken as shipped due to syntax errors.

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

Warning
Location
scripts/ask_free_ai.py:45
Finding

Automatic Cloud Fallback May Disclose Sensitive Prompt Content

Content
View full analysis
Optional[str]: """调用 Cloudflare Workers AI""" try: import requests account_id = os.getenv("CF_ACCOUNT_ID") api_token = os.getenv("CF_API_TOKEN") if not account_id or not api_token: print("[Cloudflare] 未配置", file=sys.stderr) return None url = f"https://api.cloudflare.com/client/v4/accounts/{account_id}/ai/run/@cf/meta/llama-3.1-8b-instruct" headers = {"Authorization": f"Bearer {api_token}"} data = {"prompt": prompt} resp = requests.post(url, json=data, headers=headers, timeout=60) if resp.status_code == 200: return resp.json().get("result", {}).get("response", "").strip() except Exception as e: print(f"[Cloudflare] 失败: {e}", file=sys.stderr) return None def call_groq(prompt: str) -> Optional[str]: """调用 Groq 免费 API""" try: import requests api_key = os.getenv("GROQ_API_KEY") if not api_key: print("[Groq] 未配置", file=sys.stderr) return None url = "https://api.groq.com/openai/v1/chat/completions" headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"} data = { "model": "llama-3.1-8b-instant", "messages": [{"role": "user", "content": prompt}], "max_tokens": 512 } resp = requests.post(url, json=data, headers=headers, timeout=60) if resp.status_code == 200: return resp.json().get("choices", [{}])[0].get("message", {}).get("content", "").strip() except Exception as e: print(f"[Groq] 失败: {e}", file=sys.stderr) return None def ...[truncated 3729 chars]
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (11)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The README advertises automatic routing and failover across local and cloud AI providers, but it does not clearly warn that user prompts may be transmitted to third-party services such as Cloudflare or Groq. This creates a real privacy and data-handling risk because users may reasonably assume queries stay local, especially given the emphasis on 'free' and local model support.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
94% confidence
Finding

The skill advertises and documents behavior that uses environment variables and network access to contact local and third-party AI providers, but it does not declare any explicit tool scope such as permissions or allowed-tools. This creates a transparency and control gap: users and hosts cannot easily tell that prompts and secrets may be used with external services, increasing the chance of unintended data exposure or policy bypass.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The documentation promotes automatic routing to Cloudflare Workers AI, Groq, and other cloud providers but does not warn users that their prompts may be transmitted to third parties. This is dangerous because users may submit sensitive data under the assumption the interaction is local-first, while failover or routing logic could silently send that data to external services with different retention and privacy practices.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
83% confidence
Finding

The script's natural-language descriptions, help text, status messages, and failure message are presented in Chinese, with no option to select another language. This can violate language/locale policy when a skill imposes a specific language without user opt-in or a documented region-specific justification.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
80% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · scripts/ask_free_ai.py (reported line 30)May include surrounding context.

python
"prompt": prompt,
            "stream": False
        }
        resp = requests.post(url, json=data, timeout=120)
        if resp.status_code == 200:
            return resp.json().get("response", "").strip()
    except Exception as e:

External Transmission

Medium
Category
Data Exfiltration
Confidence
90% confidence
Finding

The hardcoded Cloudflare API endpoint indicates intentional transmission of prompt data to a third-party service. By itself the URL is not the flaw, but in context it confirms an external data flow that is not clearly disclosed to the user and can undermine privacy expectations.

Content

Scanner excerpt · scripts/ask_free_ai.py (reported line 48)May include surrounding context.

python
print("[Cloudflare] 未配置", file=sys.stderr)
            return None
        
        url = f"https://api.cloudflare.com/client/v4/accounts/{account_id/ai/run/@cf/meta/llama-3.1-8b-instruct"
        headers = {"Authorization": f"Bearer {api_token}"}
        data = {"prompt": prompt}

External Transmission

Medium
Category
Data Exfiltration
Confidence
96% confidence
Finding

This request sends the full prompt to Cloudflare's external API, which is a genuine data transmission vulnerability when performed without explicit disclosure or user approval. If users paste secrets, credentials, or proprietary text, that data leaves the local environment and is exposed to third-party handling and retention policies.

Content

Scanner excerpt · scripts/ask_free_ai.py (reported line 52)May include surrounding context.

python
headers = {"Authorization": f"Bearer {api_token}"}
        data = {"prompt": prompt}
        
        resp = requests.post(url, json=data, headers=headers, timeout=60)
        if resp.status_code == 200:
            return resp.json().get("result", {}).get("response", "").strip()
    except Exception as e:

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The script sends arbitrary user prompts to Cloudflare Workers AI and Groq without any explicit notice, consent, or opt-in at the point of transmission. In a skill context, prompts may contain sensitive data, and silent routing to third-party providers creates a real confidentiality and privacy risk even if the behavior is functionally intended.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
90% confidence
Finding

The Groq API endpoint similarly evidences third-party prompt transmission. In combination with automatic routing and failover, this makes it easy for user content to be sent off-device without a clear affirmative choice, increasing privacy risk.

Content

Scanner excerpt · scripts/ask_free_ai.py (reported line 69)May include surrounding context.

python
print("[Groq] 未配置", file=sys.stderr)
            return None
        
        url = "https://api.groq.com/openai/v1/chat/completions"
        headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
        data = {
            "model": "llama-3.1-8b-instant",

External Transmission

Medium
Category
Data Exfiltration
Confidence
96% confidence
Finding

This request transmits user prompt content to Groq's hosted API, creating the same confidentiality issue as other cloud providers. In an agent skill, users may assume their content is processed locally or may not realize failover can silently send data to third parties.

Content

Scanner excerpt · scripts/ask_free_ai.py (reported line 77)May include surrounding context.

python
"max_tokens": 512
        }
        
        resp = requests.post(url, json=data, headers=headers, timeout=60)
        if resp.status_code == 200:
            return resp.json().get("choices", [{}])[0].get("message", {}).get("content", "").strip()
    except Exception as e:

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
86% confidence
Finding

The visible natural-language description and examples are entirely in Chinese, including the sample prompt "你好", and the table labels qwen2.5 as optimized for Chinese tasks. There is no statement that the skill supports multiple languages or that Chinese-only behavior is an intentional, justified locale restriction, so this may violate language/locale choice expectations.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.