mirror of
https://github.com/openai/codex.git
synced 2026-11-05 20:10:33 +00:00
jif
48524f337a
## Why Agent messages that arrive after a model request starts must survive compaction until a normal inference request includes them. Compactor output alone does not establish that those messages were delivered. ## What changed - Persist history sequences and sampling checkpoints in harness metadata, capturing the input prefix for each normal inference request. - Retain pending agent messages within existing size budgets during local and remote compaction, placing the summary before pending messages and subsequent caller input. - Preserve sampling progress across history replacement and resume, and drain queued mail after interrupted sampling even when compaction leaves context above the token limit. ## Testing Add regression coverage for request-boundary sampling checkpoints, pending messages across local compaction and repeated remote compaction with resume, and injected messages that interrupt sampling before mid-turn compaction. GitOrigin-RevId: 26c04ab96b5a1c3cc1d6625ccc74a0f00201d14c
16 lines
542 B
Rust
16 lines
542 B
Rust
mod history;
|
|
mod image_availability;
|
|
mod normalize;
|
|
mod sampling;
|
|
pub(crate) mod updates;
|
|
|
|
pub(crate) use history::ContextManager;
|
|
pub(crate) use history::HistoryReplacement;
|
|
pub(crate) use history::estimate_image_reference_bytes;
|
|
pub(crate) use history::estimate_item_token_count;
|
|
pub(crate) use history::estimate_item_token_count_with_audio;
|
|
pub(crate) use history::is_user_turn_boundary;
|
|
pub(crate) use history::rejected_audio;
|
|
pub(crate) use image_availability::replace_unresolvable_image_files;
|
|
pub(crate) use sampling::sampled_through;
|