The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or unauthorized data access behaviors were detected. The skill's architecture is well-documented and its operations are consistent with its stated purpose of report enrichment.
[COMMAND_EXECUTION]: The skill executes a local Python script (scripts/generate-enriched-report.py) to perform deterministic tasks such as infographic injection, content validation, and flipbook layout transformation. It also references standard document processing utilities like pandoc, weasyprint, and mermaid-cli as optional tools for format export.
[EXTERNAL_DOWNLOADS]: The generated HTML output references external libraries and assets from well-known services, including Chart.js from JSDelivr and fonts from Google Fonts. These are legitimate resources for high-quality web deliverables and are documented neutrally as they do not involve the execution of untrusted code within the agent's runtime environment.
[PROMPT_INJECTION]: The skill processes potentially untrusted markdown content through a sub-agent. This creates an indirect prompt injection surface.
Ingestion points: User-provided markdown files (source_path).
Boundary markers: Not explicitly specified in the sub-agent dispatch context.
Capability inventory: Execution of a local Python script, writing to project directories, and usage of Browser MCP for quality reviews.
Sanitization: Standard HTML escaping is applied for rendering, though no specific prompt-injection filtering for the input data is described.

enrich-report