Unverified50% confidenceEventExact time
A short prompt of only about ten words — 'summarize this and call the report tool' — carried six tool definitions and a JSON schema flag, causing it to be routed to the cheapest model and fall into a loop.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedWhen tool descriptions fill the context window, models are more prone to selecting the wrong tool among many similar ones.61% similarUnverifiedThe most common model failure is returning something that is 'almost JSON,' such as extra commas, markdown code fences, verbose preambles, or writing the number 3 as the word 'three,' with the AI node reporting success while the workflow crashes only several steps later when parsing fails.60% similarUnverifiedIn the OpenDiscoveryTrace dataset, each execution step is broken down and recorded across nine fields, including Thoughts, Tool Calls, Observations, Errors, Revision Triggers, and Self-reported Confidence.60% similarUnverified工具调用的可靠性高度依赖模型输出严格遵循预定义的 JSON Schema，任何字段类型偏差、缺失或多余都可能导致框架层的解析失败58% similarUnverifiedmloda 0.11 changes resolution failures from a single-line conclusion to a full elimination chain, listing each candidate plugin the resolver considered and at which gate it was eliminated.58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1000567API
curl https://kongchang.com/api/v1/knowledge/claims/1000567MCP
get_claim(id=1000567)