Static analysis reported many high-risk patterns, but review shows the dominant hits are false positives from Markdown fences, inline backticks, file extensions, URLs, and documented CLI examples. No malicious intent or prompt injection was found. The skill is safe to publish with warnings because it documents user-triggered commands, remote sharing, public tunnels, screen recording, remote assets, and writable demo editors.
The skill documents Slidev remote access, password-protected presenter mode, binding options, and public tunnel usage. These are legitimate Slidev features, but misuse can expose presentation content or controls beyond the local machine.
The referenced files clearly document network exposure features. Risk depends on user execution and configuration, so this is a warning rather than evidence of malicious behavior.
The Monaco write reference explains that edited code can be saved back to actual files. This is useful for live coding, but users should back up files and review generated edits before enabling it.
The file explicitly states that changes are saved directly to files. The behavior is documented and user-facing, not hidden or automatically triggered by the skill.
The recording reference describes camera overlay and screen capture through browser recording APIs. This is expected presentation functionality, but it can capture sensitive visual information if used without review.
The file directly documents camera and screen capture behavior. There is no upload endpoint or covert collection evidence, so the concern is privacy handling during intentional use.
Static command execution alerts are Markdown examples
The analyzer flagged many inline backticks and fenced shell examples as Ruby backtick execution. Inspection shows these are documentation examples for Slidev, npm, pnpm, export, and formatting commands, not executable code inside the skill.
The cited locations are Markdown documentation blocks and tables. No script file, subprocess API, or automatic execution path was found in the reviewed context.
Static crypto and reconnaissance alerts are false positives
The high weak-crypto and reconnaissance hits map to Markdown syntax, file extensions, headings, or harmless examples rather than cryptographic code or system probing. No evidence found of hashing, cipher use, or reconnaissance logic.
Representative flagged lines are presentation documentation and Markdown examples. The skill contains reference text, not code that performs cryptography or host discovery.
Hardcoded URLs are documentation and sample endpoints
The network URL findings are links to Slidev documentation, sample remote assets, and optional PlantUML or PDF URLs. They should be reviewed before use, but they do not show data exfiltration or hidden network activity.
The cited URLs are visible in user-facing documentation and examples. No credential transmission, telemetry, or hidden request logic was identified.