Security audit
Security checks for vulnerabilities and agentic risk
This CDN skill is mostly documentation, but it asks for command/API capability and credentials while using broad triggers and weak user-control boundaries for operations that could affect a live CDN.
Review before installing. Use this skill only for CDN tasks, and do not let it run authenticated provider API calls or shell commands unless you have explicitly requested the specific action and checked the domain, zone, token permissions, and rollback impact. Prefer read-only diagnostics unless you intend to change CDN configuration or purge production cache.
The activation description is overly broad and can cause the skill to be invoked for generic coding, debugging, deployment, or automation tasks beyond CDN administration. In a skill with exec and external API capabilities, ambiguous triggering increases the risk of inappropriate use and over-privileged actions in unrelated contexts.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
curl -X POST "https://api.cloudflare.com/client/v4/zones/{zone_id}/purge_cache"
-H "Authorization: Bearer {token}"
-d '{"files":["https://example.com/page.html"]}'
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
curl -X POST "https://api.cloudflare.com/client/v4/zones/{zone_id}/purge_cache"
-H "Authorization: Bearer {token}"
-d '{"files":["https://example.com/page.html"]}'
The documentation presents a concrete success payload with fields like configured status, cache hit rate, edge node counts, and active rule counts, implying the skill actually performs or verifies CDN changes. In this file, there is no implementation logic that would produce those operational results, creating a misleading intent-versus-behavior description.
The skill is presented as a CDN configuration assistant, but its documented capabilities expand to generic file handling, external API integration, and command execution. This capability sprawl can mislead users and downstream agents into granting broader trust and permissions than warranted, increasing the chance of unsafe execution or unintended system changes.
The skill advertises command execution, file handling, and external API use without prominent safety constraints, approval boundaries, or data handling warnings. In context, this makes the skill more dangerous because users may treat it as a normal CDN assistant while it can potentially perform side-effecting operations on the host or transmit data externally.
The manifest does not explicitly state that the skill will manage or require external provider credentials, yet the documentation instructs users to export an API key via environment variables. For a CDN advisory/configuration skill, credential handling may be valid only if provider API integration is clearly declared as part of the purpose.
No suspicious patterns detected.