The Agent Skills Directory

[DATA_EXFILTRATION]: The skill documentation describes reading the user's conversation history and memory files to inform its generation logic. It further specifies that shell scripts are used to make external API calls for rendering and asset fetching, creating a pathway for internal agent data to be transmitted to external services.
[PERSISTENCE]: The skill includes a "daily cron" runtime mode and functionality to "create a recurring job." This allows the skill to maintain an automated presence and execute at scheduled intervals within the agent's environment.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted conversation history as the primary source for its creative synthesis pipeline.
Ingestion points: Conversation context and memory files stored in workspace/daily-gift/.
Boundary markers: No delimiters or instructions to ignore embedded commands in the ingested history are specified.
Capability inventory: The skill uses shell scripts for network API calls and has the ability to read and write files in the local workspace.
Sanitization: No sanitization, validation, or filtering of the ingested context is described in the workflow.
[DYNAMIC_EXECUTION]: The skill describes a process of generating interactive H5 gifts using p5.js and canvas. This involves the assembly and execution of code based on synthesized concepts and templates at runtime.
[NO_CODE]: The provided file is a catalog entry and documentation manual; it explicitly states that it does not contain the scripts, creative seed libraries, or rendering runtime required for the described functionality.

daily-gift