bamboo-compression 2026.4.26

Compression utilities for Bamboo sessions and memory workflows
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38

//! Token budget management for LLM conversations.
//!
//! This module provides token budget management to prevent context window overflow
//! while maintaining conversation continuity. It implements a hybrid approach
//! combining rolling summary with recent message windowing.
//!
//! # Key Components
//!
//! - [`types`]: Core types like `TokenBudget`, `BudgetStrategy`, `PreparedContext`
//! - [`counter`]: Token counting via BPE tokenizer (tiktoken) or heuristic estimation
//! - [`segmenter`]: Message segmentation preserving tool-call atomicity
//! - [`limits`]: Model context window limits registry
//! - [`preparation`]: Context preparation with budget enforcement
//! - [`summarizer`]: Conversation summarization for context preservation

pub mod compression_tooling;
pub mod counter;
pub mod limits;
pub mod preparation;
pub mod segmenter;
pub mod summarizer;
pub mod types;

pub use compression_tooling::{
    active_messages_for_budget, apply_compression_plan, build_compression_plan_with_summary,
    build_forced_compression_plan_with_summary, build_summary_prompt, compression_summary_message,
    context_window_usage_percent, estimate_context_compression_exposure,
    normalized_trigger_percent, summary_source_messages, CompressionPlan, CompressionPlanError,
    ContextCompressionExposure,
};
pub use counter::{HeuristicTokenCounter, TiktokenTokenCounter, TokenCounter};
pub use limits::{create_budget_for_model, ModelLimitsRegistry};
pub use preparation::{estimate_prompt_cache_savings, prepare_hybrid_context};
pub use segmenter::MessageSegmenter;
pub use summarizer::{
    HeuristicSummarizer, LlmSummarizer, Summarizer, SummaryManager, SummaryMode, SummaryTrigger,
};
pub use types::{BudgetError, BudgetStrategy, PreparedContext, TokenBudget, TokenUsageBreakdown};