The SubagentStop hook's description now documents the JSON input format: "Input to command is JSON with agent_id, agent_type, and agent_transcript_path." This was previously undocumented, making it clearer for hook authors what data is available when a subagent completes.
Evidence
SubagentStop description update (search for "agent_id, agent_type, and agent_transcript_path")
Strings lifted out of the shipped bundle, so the claim above can be checked against them.