Unverified50% confidenceFactExact time
Some responses carry metadata related to safety filtering; when the prompt touches on sensitive topics, `text` may be empty, and you need to check the filtering reason rather than directly concluding that the call failed.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedEmails involving sensitive information or important customers should have filtering rules set up to avoid accidental triggering.75% similarUnverifiedSafety is not an inherent property of the reply text, but a relational property between the reply and the recipient's background.73% similarUnverifiedWhen a response is blocked by a content filter, Cline now explicitly informs the user that "content was blocked" and suggests rephrasing, rather than displaying "model returned an empty response."73% similarUnverifiedOutput Guardrails act after the final response is generated, reviewing the already-produced response, and can intercept non-compliant content before it is returned to the user.72% similarUnverifiedCommercial models typically have built-in safety filter layers that return refusal responses to certain types of requests; this mechanism can be removed by fine-tuning on unconstrained data.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1026477API
curl https://kongchang.com/api/v1/knowledge/claims/1026477MCP
get_claim(id=1026477)