{"repo":"NousResearch/hermes-agent","public_id":"81481","display_label":"Kimi reasoning token double-counting and compression bugs","summary":"These items address two related issues with Kimi model handling: reasoning/reasoning_content fields being charged twice in token estimates, and chat histories incorrectly triggering premature compression. The bug report documents the combined ~2.5x estimate inflation and missing replay of reasoning content on the wire.","subsystem":"agent","counts":{"members":4,"open_issues":0,"open_issues_distinct":0,"open_prs":0,"closed_issues":1,"merged_prs":1,"attempt_prs":2},"landed_fix":{"status":"merged_to_main","basis":"structural","pr":98998,"merged_at":"2026-08-31T03:40:44+00:00"},"resolution":{"state":"unresolved","by":null},"freshness":{"state":"verified","last_verified_at":"2026-08-31T03:40:44+00:00","is_stale":false},"grouping":{"backbone":"STRUCTURAL_SPANNING","link_density":0.5},"fix_edges":[{"issue_number":80246,"pr_number":98998,"tier":"verified","basis":"superseded_ref","confidence":0.9,"recoverable_from_timeline":false,"pr_merged_at":"2026-08-31T03:40:44+00:00"},{"issue_number":81481,"pr_number":98998,"tier":"verified","basis":"superseded_ref","confidence":0.9,"recoverable_from_timeline":false,"pr_merged_at":"2026-08-31T03:40:44+00:00"},{"issue_number":84371,"pr_number":98998,"tier":"verified","basis":"closing_ref","confidence":0.95,"recoverable_from_timeline":true,"pr_merged_at":"2026-08-31T03:40:44+00:00"}],"members":[{"kind":"issue","number":81481,"title":"[Bug]: reasoning/reasoning_content double-charged on every historical turn, neither replayed on kimi chat/completions wire — residual ~2.5x estimate inflation after #75884","url":null,"state":"closed"},{"kind":"pr","number":81504,"title":"fix(compression): stop Kimi histories from triggering premature compression","url":null,"state":"closed","close_reason":{"refs":[{"kind":"commit","value":"26f178e5fa","number":0},{"kind":"commit","value":"452f6b7de2","number":0},{"kind":"commit","value":"64cc87e668","number":0},{"kind":"pr","value":"#98998","number":98998},{"kind":"pr","value":"#81481","number":81481},{"kind":"pr","value":"#92690","number":92690}],"category":"implemented_elsewhere","implication":"superseded"},"merged":false},{"kind":"pr","number":81509,"title":"fix(estimator): charge duplicate reasoning fields once in token estimates","url":null,"state":"closed","close_reason":{"refs":[{"kind":"pr","value":"#81504","number":81504}],"category":"superseded_by_pr","implication":"superseded"},"merged":false},{"kind":"pr","number":98998,"title":"fix(compression): compaction dead-loop ended — trigger and tail walks charge stale thinking by wire truth (#84371)","url":null,"state":"merged","close_reason":{"category":"merged"},"merged":true,"is_fix_pr":true,"fix_tier":"verified","files_changed":["agent/context_compressor.py","agent/conversation_compression.py","agent/conversation_loop.py","agent/message_sanitization.py","agent/model_metadata.py","agent/turn_context.py","tests/agent/test_estimator_parity_84371.py"]}],"trust":{"default_tier":"verified","includes_suggested":false}}