Security audit
Security checks for vulnerabilities and agentic risk
This logo skill is not clearly malicious, but it asks for broad read/write/command authority while its documentation mixes logo generation with unrelated automation, code, ops, and API capabilities.
Review before installing. Use this only if you are comfortable granting a logo skill broad read, write, and command-execution authority, and avoid invoking it for operations, code, deployment, or bulk file tasks unless the publisher narrows and documents those capabilities.
The skill is presented as a logo-generation tool, but this section advertises unrelated code analysis, dependency scanning, and CI/CD capabilities. That mismatch can mislead an agent or user into granting broader trust and invoking the skill in contexts far beyond its stated purpose, increasing the risk of unintended execution paths and over-privileged use.
The documented contract is generic content-processing output with a 'reviewer' template rather than a logo-design interface. This inconsistency makes it unclear what the skill really does and can cause downstream agents to pass arbitrary content or interpret outputs incorrectly, which is dangerous when the skill also declares write and exec tooling.
This section expands the skill from logo design into file handling, external API use, and system command execution. Broad operational claims combined with tool access create a materially larger attack surface because an agent may authorize sensitive filesystem or shell actions under the guise of a creative design skill.
The manifest description mixes logo design with operations monitoring, log analysis, alerting, and deployment management. This is a severe scope-confusion issue: the skill’s identity no longer matches its use cases, so agents may invoke it in sensitive operational contexts where a creative skill should never apply.
These sections redefine the skill as a generic automation and execution tool rather than a narrowly scoped logo assistant. In context, that broadening is especially dangerous because the skill declares read/write/exec tools, so ambiguous positioning can normalize use in high-risk tasks without adequate scrutiny.
The invocation description is overly broad and ambiguous, spanning multiple unrelated domains. Ambiguous trigger scope increases the chance that an agent auto-selects this skill for inappropriate tasks, which is risky given the presence of powerful tools and unclear operational boundaries.
The quick-start tells users to invoke the skill with 'necessary input parameters' but does not define a constrained trigger scope or safe usage boundaries. That vagueness can lead to arbitrary or opportunistic use, especially when the rest of the document contains conflicting generic automation language.
The markdown advertises file writing, API integration, and command execution but does not prominently warn users about the security implications. In a skill that appears creative and low-risk on its face, undocumented operational side effects can cause users or agents to underestimate the sensitivity of the granted capabilities.
No suspicious patterns detected.
