The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted user input via the flowData parameter in SKILL.md, which creates a surface for indirect prompt injection attacks.
Ingestion points: The nodes.label and edges.label properties within the flowData object are primary ingestion points for external strings.
Boundary markers: The skill definition lacks explicit boundary markers or instructions to the agent to disregard embedded commands within these labels.
Capability inventory: The skill is configured with Bash, Read, Write, and Edit tools; if the agent were to obey instructions injected into labels, it could perform file operations or command execution.
Sanitization: There is no evidence of sanitization, validation, or escaping logic for the user-supplied text in the provided documentation.
[EXTERNAL_DOWNLOADS]: The skill integrates with the diagrams.net API for export functionality and utilizes Mermaid.js and PlantUML for diagram generation. These are well-known services and libraries appropriate for the skill's documented purpose.
[NO_CODE]: The skill is composed entirely of markdown documentation and schema definitions, with no included executable scripts or source code files beyond the instructions themselves.

user-flow-diagram