Shared from "Study" on Inkdown
The context management system prevents token overflow while preserving critical information. It intelligently trims conversation history using multiple strategies based on token costs.
File: src/server/repositories/engine/engine.ts:201
File: src/server/repositories/engine/engine.ts:38
Layout Options:
How It Works:
File: src/server/repositories/engine/engine.ts:148
File: src/server/repositories/engine/engine.ts:133
Returns messages unchanged.
File: src/server/repositories/engine/engine.ts:94
Uses tool-provided summary if shouldIncludeInHistory: false.
File: src/server/repositories/engine/engine.ts:46
Uses LLM to generate summary of conversation history.
File: src/server/repositories/engine/engine.ts:142
File: src/server/endpoints/unified/features/deepResearch/memory/index.ts
For deep research, maintains research context across steps:
Why Context Scoring:
Different plans get different context limits:
The context management system:
Key Principle: Preserve maximum useful information while staying within token limits. Summarize only when necessary, use tool summaries when available.