{"repo":"NousResearch/hermes-agent","public_id":"116078","display_label":"VRAM-based resident model cap for local runtime","summary":"A bug was identified where the local runtime admitted resident models by count rather than available VRAM, causing silent paging and slowdowns on limited-memory cards. The fixes derive the residency cap from actual VRAM and staged model sizes rather than a simple count ceiling.","subsystem":null,"counts":{"members":3,"open_issues":0,"open_issues_distinct":0,"open_prs":0,"closed_issues":1,"merged_prs":1,"attempt_prs":1},"landed_fix":{"status":"merged_to_main","basis":"structural","pr":117656,"merged_at":"2026-09-21T03:42:06+00:00"},"resolution":{"state":"resolved","by":"pr-117656"},"freshness":{"state":"verified","last_verified_at":"2026-09-21T03:42:06+00:00","is_stale":false},"grouping":{"backbone":"STRUCTURAL_SPANNING","link_density":0.6666666666666666},"fix_edges":[{"issue_number":116078,"pr_number":117656,"tier":"verified","basis":"closing_ref","confidence":0.95,"recoverable_from_timeline":true,"pr_merged_at":"2026-09-21T03:42:06+00:00"}],"members":[{"kind":"issue","number":116078,"title":"[Bug]: Managed local runtime admits models by count, not by VRAM — on an 8 GB card the second resident is silently paged (2.6× slower)","url":null,"state":"closed"},{"kind":"pr","number":116350,"title":"fix(local-models): derive the residency cap from the card and the staged models, not the count alone","url":null,"state":"closed","close_reason":{"refs":[{"kind":"commit","value":"1142a5b3729e","number":0},{"kind":"pr","value":"#117656","number":117656}],"category":"merged_via","implication":"superseded"},"merged":false},{"kind":"pr","number":117656,"title":"fix(local-runtime): resident model cap derives from VRAM with models_max as the ceiling (#116078, salvage #116350)","url":null,"state":"merged","close_reason":{"category":"merged"},"merged":true,"is_fix_pr":true,"fix_tier":"verified","files_changed":["hermes_cli/local_runtime/bootstrap.py","hermes_cli/local_runtime/presets.py","tests/hermes_cli/test_local_residency_admission.py"]}],"trust":{"default_tier":"verified","includes_suggested":false}}