The Respan Cursor integration captures AI agent conversations in real-time, providing visibility into thinking blocks, tool calls, file edits, and shell commands through hierarchical traces.
Hooks fire during agent execution and capture events including thinking blocks, file edits, shell commands, and MCP tool calls as child spans beneath a root agent span. Traces include unique identifiers linking turns within conversation threads.
Set KEYWORDSAI_API_KEY and TRACE_TO_KEYWORDSAI=true as environment variables in your shell profile.
Download the Python hook script to ~/.cursor/hooks/ and create a hooks.json configuration with seven hook triggers for prompt submission, agent thoughts, shell execution, file edits, MCP execution, agent responses, and stop events.
Restart Cursor IDE. Traces appear hierarchically in the Respan dashboard with a root agent span and child spans for each event type.