{"repo":"NousResearch/hermes-agent","public_id":"104282","display_label":"Video input handling bugs in Codex adapter","summary":"These issues concern a bug where video input passed through the Codex Responses adapter is silently dropped or mishandled, resulting in text-only responses or ungrounded output instead of proper errors. The members address detection and correction of this silent failure mode.","subsystem":"agent/transports","counts":{"members":3,"open_issues":0,"open_issues_distinct":0,"open_prs":1,"closed_issues":1,"merged_prs":1,"attempt_prs":1},"landed_fix":{"status":"merged_to_main","basis":"structural","pr":104953,"merged_at":"2026-09-07T13:04:45+00:00"},"resolution":{"state":"resolved","by":"pr-104953"},"freshness":{"state":"verified","last_verified_at":"2026-09-07T13:04:45+00:00","is_stale":false},"grouping":{"backbone":"STRUCTURAL_SPANNING","link_density":0.6666666666666666},"fix_edges":[{"issue_number":104282,"pr_number":104953,"tier":"verified","basis":"closing_ref","confidence":0.95,"recoverable_from_timeline":true,"pr_merged_at":"2026-09-07T13:04:45+00:00"}],"members":[{"kind":"issue","number":104282,"title":"[Bug]: video_analyze silently drops video input when routed through the Codex Responses adapter","url":null,"state":"closed"},{"kind":"pr","number":104292,"title":"fix(vision): stop Codex video analysis from returning ungrounded text","url":null,"state":"open","close_reason":{"category":"uncategorized","implication":"needs_llm"},"merged":false},{"kind":"pr","number":104953,"title":"fix: unsupported video input no longer becomes text-only success","url":null,"state":"merged","close_reason":{"category":"merged"},"merged":true,"is_fix_pr":true,"fix_tier":"verified","files_changed":["agent/codex_responses_adapter.py","evals/provider_wire/issue_104282.py","tests/agent/test_auxiliary_client.py","tests/agent/test_codex_responses_adapter.py"]}],"trust":{"default_tier":"verified","includes_suggested":false}}