Back to skill

Security audit

LLM Gateway — Multi-Model AI Completions

Security checks for vulnerabilities and agentic risk

Overview

This is a simple LLM gateway skill, but it sends prompts in URL paths and can trigger automatic USDC payments without documented limits.

Review this skill before installing. Use it only with non-sensitive prompts unless you have separate privacy assurances from the provider, and configure wallet controls such as low balances, per-transaction approval, exact price and recipient checks, and task-level spending limits.

Vulnerability Patterns
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
  • 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
Findings (2)

T09 · Insecure Skill Coding Practices

Warning
Location
SKILL.md:18
Finding

Sensitive Prompt Disclosure Through URL Paths

Content
View full analysis
Remediation
View remediation

T05 · Unauthorized Access and Privilege Escalation

Warning
Location
SKILL.md:18
Finding

Unbounded Automatic USDC Payment Workflow

Content
View full analysis
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • 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 (4)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill instructs agents to place the full prompt directly in the URL path and send it to a third-party service, but it does not warn that prompts may contain sensitive user or system data. Putting prompts in the URL is especially risky because URLs are commonly logged by clients, proxies, gateways, browser/history layers, and server infrastructure, increasing exposure beyond normal request-body transmission.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
94% confidence
Finding

This line documents sending user prompts to an external endpoint, which is a genuine data exfiltration boundary for any agent using the skill. Because the prompt is embedded in the URL path, the exposure surface is broader than a typical API call and may leak task contents into network and application logs.

Content

Scanner excerpt · SKILL.md (reported line 24)May include surrounding context.

md
## Tiers (live endpoints)
| Call | Model tier | Price |
|---|---|---|
| `GET https://api.gocreativeai.com/v1/ai/ask/{prompt}` | Fast & cheap (DeepSeek-tier) | ~$0.02 |
| `GET https://api.gocreativeai.com/v1/ai/pro/{prompt}` | Frontier — Claude 4.5 Sonnet (`?model=` to override) | ~$0.08 |
| `GET https://api.gocreativeai.com/v1/ai/ultra/{prompt}` | Ultra — Claude Opus 4.6, top-end reasoning | ~$0.25 |

External Transmission

Medium
Category
Data Exfiltration
Confidence
94% confidence
Finding

This endpoint similarly transmits prompt content to a third-party LLM gateway and therefore creates an external data-sharing risk. The skill context makes this more dangerous because agents may forward arbitrary user inputs, internal instructions, or confidential workspace data to the provider without a strong warning or consent flow.

Content

Scanner excerpt · SKILL.md (reported line 25)May include surrounding context.

md
| Call | Model tier | Price |
|---|---|---|
| `GET https://api.gocreativeai.com/v1/ai/ask/{prompt}` | Fast & cheap (DeepSeek-tier) | ~$0.02 |
| `GET https://api.gocreativeai.com/v1/ai/pro/{prompt}` | Frontier — Claude 4.5 Sonnet (`?model=` to override) | ~$0.08 |
| `GET https://api.gocreativeai.com/v1/ai/ultra/{prompt}` | Ultra — Claude Opus 4.6, top-end reasoning | ~$0.25 |

URL-encode the prompt. Example: `/v1/ai/ask/Summarize%20this%20contract%20in%203%20bullets`.

External Transmission

Medium
Category
Data Exfiltration
Confidence
94% confidence
Finding

The ultra-tier endpoint is another explicit external transmission path for prompt contents to a remote service. The higher-cost reasoning use case may encourage users to send larger, more sensitive problem statements, increasing the chance that confidential information is disclosed through both provider access and URL logging.

Content

Scanner excerpt · SKILL.md (reported line 26)May include surrounding context.

md
|---|---|---|
| `GET https://api.gocreativeai.com/v1/ai/ask/{prompt}` | Fast & cheap (DeepSeek-tier) | ~$0.02 |
| `GET https://api.gocreativeai.com/v1/ai/pro/{prompt}` | Frontier — Claude 4.5 Sonnet (`?model=` to override) | ~$0.08 |
| `GET https://api.gocreativeai.com/v1/ai/ultra/{prompt}` | Ultra — Claude Opus 4.6, top-end reasoning | ~$0.25 |

URL-encode the prompt. Example: `/v1/ai/ask/Summarize%20this%20contract%20in%203%20bullets`.