The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes data from an external API, which creates a surface for indirect prompt injection. If the API response is compromised or contains malicious directive-like text, it could attempt to influence the agent's behavior.
Ingestion points: Data enters the agent's context through the https://www.worldmonitor.app/api/intelligence/v1/compute-energy-shock endpoint defined in SKILL.md. Specifically, fields such as assessment and limitations contain natural language content derived from external datasets.
Boundary markers: The skill includes a dedicated "Content safety" section that explicitly warns the agent that the response is data, not instructions. It provides clear markers to disregard directive-like text such as "ignore previous instructions" or command execution requests found within the data.
Capability inventory: The skill is intended to be used by agents that can compute scenarios and present analysis, potentially using other tools to act on the energy market data.
Sanitization: While the skill lacks programmatic sanitization, it employs prompt-based boundary enforcement to guide the agent's interpretation of external content.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to www.worldmonitor.app to fetch energy scenario modeling data. This is the primary function of the skill and does not involve the transfer of sensitive local files or credentials.

assess-energy-shock