In Claude Code's agentic loop, what triggers a new API call after a to — Claude Code Certified Professional Practice — Claude Cert Academy
In Claude Code's agentic loop, what triggers a new API call after a tool executes?
- Claude automatically retries the original prompt with the tool result appended
- The tool result is returned as a tool_result message and Claude generates the next response in a new API call
- The user must manually press Enter to send the tool result to Claude
- The tool result is cached locally and Claude references it in its next response without an API call
Answer: The tool result is returned as a tool_result message and Claude generates the next response in a new API call
After tool execution, the result is added as a tool_result block and a new API call is made. Claude processes the result and generates its next response — this cycle is the agentic loop.
Continue to Claude Cert Academy