Unverified50% confidenceOpinionExact time
Context is not a chat history but a workbench assembled before execution; the issue lies not in window length but in information density.
1
Sources
50%
Confidence
Long-term
Relevance
9/19/2026
First Seen
Sources
Related Entities
Related Claims
Unverified将工作拆分为独立工单可缓解LLM对话过长时的上下文遗忘问题，使每个对话的上下文窗口保持精简74% similarUnverifiedIn large model usage, the context window is a limited resource, and token consumption accumulates with conversation turns, file reads, and code injection.71% similarUnverifiedA single Tab completion in Cursor may consume thousands of input tokens for context construction, whereas a pure conversation scenario does not incur this hidden cost68% similarUnverifiedChat mode only answers questions without executing, Work mode completes office tasks end-to-end, and Codex mode completes development tasks within code repositories.67% similarUnverifiedSplitting work on different topics or phases into separate sessions results in shorter conversation pages that are more friendly to lazy loading and screenshot tools, reducing the technical difficulty of individual exports.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/934430API
curl https://kongchang.com/api/v1/knowledge/claims/934430MCP
get_claim(id=934430)