The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by design, as it handles external prompt data intended for LLM processing.
Ingestion points: The messages and prompts arguments in the balanced_completion and parallel_completions functions within SKILL.md directly ingest untrusted input.
Boundary markers: The provided code examples do not implement delimiters or explicit instructions for the agent to ignore embedded commands within the processed data.
Capability inventory: The skill utilizes the requests library and OpenAI SDK to perform network operations to openrouter.ai.
Sanitization: Input data is passed to the API without prior sanitization or validation.

openrouter-load-balancing