The Agent Skills Directory

[SAFE]: The skill instructions focus on logical reasoning, trade-off analysis, and documentation verification. There are no attempts to bypass safety filters, exfiltrate data, or execute unauthorized commands.
[INDIRECT_PROMPT_INJECTION]: The skill processes user requests for brainstorming. It includes safety-oriented instructions to verify capabilities using external documentation tools (context7 MCP) rather than relying on potentially poisoned context or memory, which reduces the surface for indirect injection via library assumptions.
[COMMAND_EXECUTION]: The skill references specific roles and tools (ork_explorer, context7), which are used as part of its intended functionality for information retrieval and synthesis within a controlled environment.

ork-brainstorm