AgentRXI Context Packer

Choose this endpoint when you need to fit relevant information from long documents or web pages into a constrained token window for an LLM prompt. It is especially useful in RAG pipelines, multi-source summarization, or when you want to reduce cost by sending only the relevant parts of large texts rather than the full content. Prefer it over generic summarization when you have a specific question and need source-attributed passages rather than a freeform summary.

• URL is inaccessible or returns non-200 — passages from that source will be missing or an error noted
• Token budget too small to include any meaningful passage — may return minimal or empty context
• Question is too vague — relevance ranking may return low-quality or off-topic passages
• Input text or URL content is not in English or is non-textual (e.g. PDF with images) — extraction may fail or return garbage
• Rate limiting or payment failure — returns 402 or service error