The Agent Skills Directory

[SAFE]: No evidence of malicious behavior, obfuscation, or unauthorized data access was found. The skill operates as a persona-based prompt for analytical and creative interaction.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input to generate persona-based responses, creating a potential surface for prompt injection.
Ingestion points: User queries are ingested in SKILL.md through the "Agentic Protocol" workflow.
Boundary markers: None are present to distinguish user input from instructions.
Capability inventory: The skill instructions in SKILL.md involve reading local files references/research.md and examples/demo-conversation.md. No network operations, file writing, or shell command capabilities are present.
Sanitization: No input sanitization or filtering mechanisms are implemented.

borges-perspective