Unverified50% confidenceFactExact time
Claude Code's Agent loop does not rely on stop_reason to decide whether to continue; instead it directly checks whether the content array of the assistant message contains a tool_use block.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/13/2026
First Seen
Valid until: 1/11/2027
Sources
Related Entities
Related Claims
UnverifiedClaude determines the loop termination condition by checking whether stop_reason equals tool_use, while OpenAI checks whether the output contains a function call.76% similarUnverifiedClaude Code的自动检查是按方案运行的一段可重复检查，只有实际运行跑通才算数，AI声称已检查而未实际运行的不算67% similarUnverifiedClaude Code can automatically analyze Git change records and run only the affected API test cases without requiring full regression.65% similarUnverifiedExecution hooks serve as a trust-verification layer in Agent systems, checking whether a task has truly been completed through independent code logic rather than relying on the model's self-reporting.64% similarUnverifiedagent-ops-check clearly distinguishes 'maps to' from 'compliant'; the tool only does mapping, not compliance claims.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1013813API
curl https://kongchang.com/api/v1/knowledge/claims/1013813MCP
get_claim(id=1013813)