Module compress 

Structs

AiCompressor

AI-based compressor using a Provider.

CompressionBias

Compression bias - controls what to prioritize during compression.

CompressionConfig

Configuration for context compression.

CompressionHistoryEntry

Compression history entry for session metadata.

CompressionResult

Result of a compression operation.

SummarizedSegment

A segment of conversation history that has been summarized.

Enums

CompressionStrategy

Strategy for compressing conversation history.

Constants

DEFAULT_COMPRESSION_THRESHOLD

Compression trigger threshold (percentage of context window).

DEFAULT_COMPRESSOR_MODEL

Default model for summarization (cost-effective).

DEFAULT_TARGET_RATIO

Target ratio after compression (keep this fraction of tokens).

MIN_MESSAGES_TO_KEEP

Minimum messages to keep after compression.

Traits

Compressor

Compressor trait for different compression implementations.

Functions

build_summary_prompt

Build a prompt for AI-based summarization.

compress_messages

Compress messages synchronously (for non-AI strategies).

compress_messages_with_ai

Compress messages with AI summarization (async version).

compress_with_bias

Compress messages with bias - prioritized removal based on configuration.

estimate_tokens

Estimate token count for a message (improved approximation). For mixed content (code, Chinese, English), use a weighted estimate:

estimate_total_tokens

Estimate total tokens for a message list.

format_tokens

should_compress

Check if compression should be triggered.