ailoop-context

Conversation history management and compaction for ailoop.

Provides:

• ContextManager — keeps the conversation message vector and the pin mask used by compaction strategies.
• Compactor strategies — drop / summarize old turns while preserving tool-call / tool-result pairing. Emits StreamChunk::HistoryCompacted with strategy name and counts.
• HistoryStore — async trait for persisting and restoring conversation state. Ships with InMemoryHistoryStore and JsonFileHistoryStore.
• Tokenizer / CharTokenizer — pluggable token-estimation abstraction used by budgeting strategies.
• ConversationSnapshot — versioned, JSON-serializable representation of an active Conversation, ready for restart.

Most application code reaches the context surface through the ailoop façade re-exports rather than depending on this crate directly.

See the workspace README for the big picture.

License

Licensed under either of Apache-2.0 or MIT, at your option.