Back to skill

Security audit

ComfyUI Generator

Security checks for vulnerabilities and agentic risk

Overview

The skill appears to be a ComfyUI image-generation integration, but it uses a reusable default API key over HTTP and can send local images to the configured server without clear warning.

Review the configuration before installing. Use this only with a trusted local ComfyUI server or an HTTPS endpoint, replace the documented default API key with a unique secret, and avoid sending private images unless you trust the configured server. Ask the publisher to document any file monitoring or archiving behavior before enabling that feature.

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/comfy_client.py:18
Finding

Hardcoded Default Bearer Credential Transmitted over Plaintext HTTP

Content
View full analysis
Remediation
View remediation
`. - Instruct users to generate a unique, high-entropy credential for each deployment. - Store credentials in a protected secret manager or environment variable rather than source-controlled files. 3. Enforce secure transport: - Require HTTPS for all non-loopback endpoints. - Permit HTTP only for verified loopback addresses such as `127.0.0.1`, `::1`, or `localhost`. - Reject remote plaintext URLs rather than merely warning about them. 4. Attach authentication conditionally: - Do not add an `Authorization` header when no genuine credential is configured. - Ensure credentials are sent only to an explicitly trusted origin. - Avoid forwarding authorization headers across redirects to a different host. 5. Harden HTTP behavior: - Disable or carefully validate redirects for authenticated requests. - Validate the configured scheme and hostname before creating the session. - Use correctly configured TLS certificate verification for HTTPS endpoints. 6. Rotate the documented/default credential anywhere it has already been used and review exposed ComfyUI instances for unauthorized activity. ]]>
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 (3)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
86% confidence
Finding

For a markdown file, missing-warning findings apply when the description omits warnings about behaviors that could affect user data or system integrity. "File monitoring and archiving" implies automated handling of user files, but the README provides no caution about what directories may be watched, what files may be archived, or whether files may be moved or copied.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

This code reads a local image file and uploads it to a remote ComfyUI endpoint, which is a safety-relevant network transmission of user data. Although errors are logged, there is no confirmation prompt or explicit disclosure near this operation warning that local file contents will be sent to the server.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
92% confidence
Finding

This code contains natural-language strings such as the module docstring and class docstring in Chinese, while generated prompt content and some suggestions are in English. Because the skill does not provide any user opt-in or configuration for language/locale, it may violate a policy requiring language choice rather than imposing one by default.

Content

No source excerpt is available for this finding.