pretty-mermaid
Warn
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The scripts
scripts/render.mjs,scripts/batch.mjs, andscripts/themes.mjsutilizeexecSyncto programmatically executenpm installif required dependencies are missing from the environment. - [DYNAMIC_EXECUTION]: The skill uses the
import()function with a path computed at runtime to load thebeautiful-mermaidlibrary, which is a technique that can be used to execute arbitrary code if the local filesystem is compromised. - [EXTERNAL_DOWNLOADS]: The skill is designed to automatically download and install the
beautiful-mermaidpackage from the public npm registry during its first run if it is not already present. - [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: User-provided Mermaid diagram source code or files passed to the
--inputor--input-dirarguments inscripts/render.mjsandscripts/batch.mjs. - Boundary markers: None. The scripts read the raw content of the input files and pass them directly to the rendering engine.
- Capability inventory: The skill has capabilities for file system read/write operations (
readFileSync,writeFileSync,mkdirSync,readdirSync) and shell command execution (execSync). - Sanitization: No sanitization or validation of the Mermaid code is performed by the scripts before it is processed by the external library.
Audit Metadata