Function filter_reasoning_artifacts 

pub fn filter_reasoning_artifacts(history: &mut [ChatMessage]) -> usize

Mid-session model switch (§1.10, dep 8): strip every reasoning artifact out of history in place — both the metadata keys (REASONING_METADATA_KEYS) and any content_parts blocks whose "type" is in REASONING_CONTENT_PART_TYPES — so model-A’s reasoning never reaches model-B’s context on the next request built from this history. Returns the count of MESSAGES actually touched (had at least one metadata key removed and/or at least one content part removed) — Agent::switch_model records this in the persisted model_change entry (ModelChangeRecord::reasoning_artifacts_filtered).

Pure and total: never panics, a message with nothing to filter is left byte-identical (down to content_parts ordering of the parts that survive), and calling this with nothing to filter (the common case — supercode’s own live loop never populates these keys) is a cheap no-op scan, not a rebuild.