The Agent Skills Directory

[SAFE]: The skill consists entirely of Markdown instructions and reference documentation for generating Mermaid diagrams. It does not contain any executable scripts, binary files, or configuration that triggers tool usage.
[PROMPT_INJECTION]: No malicious instruction override patterns or safety bypass attempts were detected. The instructions focus purely on diagram syntax and formatting guidelines.
[DATA_EXFILTRATION]: The skill does not contain any network operations (e.g., curl, wget, fetch) or access to sensitive file paths (e.g., .ssh, .aws, .env). There is no mechanism for exfiltrating data.
[COMMAND_EXECUTION]: No shell commands, subprocess calls, or system interactions were identified. The skill's workflow is limited to generating Markdown-compatible Mermaid code.
[REMOTE_CODE_EXECUTION]: The skill does not download external scripts, install packages (npm/pip), or execute remote code. All logic is contained within the provided Markdown files.
[OBFUSCATION]: Analysis of the text and syntax rules revealed no hidden content, Base64-encoded commands, zero-width characters, or homoglyph-based evasion techniques.
[INDIRECT_PROMPT_INJECTION]: While the skill processes user-supplied text to generate diagrams, it possesses no exploitable capabilities (such as file writing or network access). The risk is safely managed as the output is restricted to plain-text diagram code.
Ingestion points: User-provided text concepts in SKILL.md workflow.
Boundary markers: The workflow instructions implicitly define the input as data to be analyzed, though explicit delimiters are not enforced.
Capability inventory: None. The skill only outputs text.
Sanitization: Includes specific rules to prevent Mermaid syntax errors, which acts as a form of output sanitization.

mermaid-visualizer