The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external web sources (e.g., style research from art galleries, design festivals, and trade publications). This creates a vulnerability surface where malicious instructions could be embedded in the content of the linked websites or the research data processed by the agent.
Ingestion points: The file references/style-intelligence-ledger.md contains extensive lists of external URLs (such as mplus.org.hk, vogue.com, and wallpaper.com) that the skill uses as "evidence tiers" for visual signals.
Boundary markers: The skill instructions lacks explicit boundary markers or instructions to ignore potential commands embedded in the external research content during interpolation.
Capability inventory: The skill is primarily focused on generating text-based art-direction intelligence and routing to companion skills; it does not demonstrate capabilities for arbitrary command execution or unauthorized file system access.
Sanitization: No sanitization or filtering mechanisms are defined for the external data before it is distilled into reusable "style controls."- [EXTERNAL_DOWNLOADS]: The skill references numerous external URLs for research purposes. These references target well-known organizations and reputable services in the design, art, and technology sectors (e.g., Apple, Google DeepMind, Figma, and various international museums). These references are documented for research context and do not involve executable code downloads.

aesthetic-style-intelligence