Security audit
Security checks for vulnerabilities and agentic risk
This skill is a straightforward skill generator that writes new skill files by design, with no evidence of hidden execution, exfiltration, or persistence beyond generated files.
Install only if you want an agent to create or rewrite skill files in your workspace. Review the generated SKILL.md, README, target directory, external services, and any helper scripts before enabling a generated skill, especially if it uses APIs, credentials, or posting actions.
The README explicitly states that the skill 'generates all files' and 'writes files' but does not warn users that invoking it can modify the local filesystem. In an agent setting, undocumented write behavior can lead to unintended file creation or overwriting, especially if the generated path or scope is not clearly constrained before execution.
The examples encourage creation of skills that call free APIs, monitor GitHub Actions, and post to Notion, but the README does not disclose that generated skills may send user-provided data, tokens, or workflow metadata to third-party services. In an agent ecosystem, this omission increases the risk of unintentional data exfiltration or unsafe handling of credentials during use of generated skills.
The usage section defines activation conditions in broad functional terms without clear exclusion criteria, making it ambiguous when this skill should be selected over normal assistance. In a meta-skill that can generate production-ready agent skills and write files, ambiguous invocation raises the risk of accidental skill execution, unnecessary file creation, and propagation of unsafe or overpowered workflows.
The trigger phrases are extremely broad and can cause this meta-skill to activate for many ordinary requests about building tools, automations, or capabilities. Because this skill creates new skills, overbroad routing increases the chance of unintended delegation, policy bypass through generated artifacts, or unsafe expansion of capabilities without explicit user intent.
The description says the skill generates agent skills from "natural language" without narrowing the trigger scope, constraints, or activation conditions. In a manifest file, this broad phrasing can make it unclear what kinds of requests should invoke the skill versus ordinary conversation.
This markdown template leaves the invocation description entirely as {{usage_triggers}} without requiring concrete trigger phrases, scope limits, or exclusion conditions. As written, downstream skills generated from this template could easily omit specificity and end up with vague activation criteria that overlap with ordinary user language.
No suspicious patterns detected.
