{"id":"ghsa-ghsa-x567-49g5-p4cf","seq":2056930,"kind":"advisory","title":"vLLM: Responses API bypasses trust_request_chat_template for chat_template_kwargs","url":"https://github.com/advisories/GHSA-x567-49g5-p4cf","permalink":"https://vulnpulse.app/items/ghsa-ghsa-x567-49g5-p4cf","summary":"影響: pip:vllm >= 0.16.0, < 0.32.0 → 0.32.0\nvLLM's /v1/responses endpoint accepts request-level chat_template_kwargs, but does not apply the --trust-request-chat-template gate that protects Chat Completions, render, tokenize, and pooling paths. When the server is run with the documented default trust_request_chat_template=False and no explicit --chat-template, a Responses request can pass: json {\"chat_template_kwargs\": {\"chat_template\": \" \"}} That key is preserved by ChatParams.get_apply_chat_template_kwargs() and then bound as the chat_…","severity":"medium","techs":["python"],"cves":["CVE-2026-115918"],"tags":["GHSA-x567-49g5-p4cf","vllm"],"source":{"id":"ghsa","name":"GitHub Advisory Database"},"published_at":"2026-10-28T21:14:59.000Z","fetched_at":"2026-10-29T00:48:52.653Z"}