The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill uses a two-stage process where user input influences an 'algorithmic philosophy' which in turn guides code generation. This architecture creates a surface where malicious instructions embedded in the user's artistic request could influence the generated p5.js code. ● Ingestion points: User artistic concepts and 'conceptual seeds' processed in the philosophy creation phase. ● Boundary markers: No explicit delimiters are used to separate the user-influenced philosophy from the code implementation logic. ● Capability inventory: The agent generates and executes p5.js code in an interactive HTML viewer. ● Sanitization: No sanitization is performed on the inferred conceptual DNA before it is woven into the algorithm.
[EXTERNAL_DOWNLOADS]: The skill's HTML viewer template fetches the p5.js library from Cloudflare's cdnjs and fonts from Google's font services. These are established, well-known providers of web assets.

algorithmic-art