Security audit
Security checks for vulnerabilities and agentic risk
This skill is a straightforward Mermaid-to-image converter that uses an external rendering service, with the main risk being that diagram text is sent to mermaid.ink.
Install only if you are comfortable with Mermaid diagram contents being sent to mermaid.ink. Do not use it for diagrams containing secrets, private infrastructure details, customer data, or other sensitive information unless you switch to a local renderer.
The skill invokes shell commands and external network access via Node.js and curl, but it does not declare any explicit tool scope or permissions boundaries. This can cause the agent to run shell-capable behavior without clear policy constraints, increasing the risk of unintended command execution or use in environments that expect least-privilege declarations.
The activation guidance is broad enough to match common requests like 'draw a diagram' or 'make a visual,' which can cause the skill to trigger in situations where the user did not intend to send content to an external service. In this skill's context, overbroad routing is more concerning because the workflow transmits diagram contents to mermaid.ink, creating accidental data exposure risk.
The documentation states the tool uses only Node.js built-in modules, but the implementation actually relies on executing curl. This misleading claim can cause operators to deploy or trust the skill under false assumptions, especially in locked-down or allowlisted environments where subprocess execution and external binaries are restricted.
The script base64-encodes the Mermaid source and sends it to https://mermaid.ink/..., which discloses the full diagram contents to a third-party service. If users supply proprietary architecture diagrams, credentials embedded in diagrams, or internal system details, this can lead to unintended data exposure; the skill context increases the risk because diagram-generation workflows often involve sensitive internal designs.
The stated purpose is diagram-to-image conversion via the mermaid.ink API, which does justify network access, but invoking an external binary is a broader capability than the manifest's core functional description suggests. The manifest and file header emphasize minimal built-in-module usage, yet the implementation depends on spawning curl, adding process-execution capability that is not clearly justified as part of the skill's intended scope.
Detected: suspicious.dangerous_exec
