When a model download proxy is configured, createProxyFetch() returned the
Response from undici's own fetch, which is not the global Response.
Transformers.js only writes `instanceof Response` results to its file cache,
so the ONNX weights were never cached and preload failed with "Unable to get
model file path or buffer." Keep undici's fetch and ProxyAgent, and re-wrap the
response body stream in a global Response.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: multica-agent <github@multica.ai>
每次模型请求前检查总预算并裁剪工具结果副本，保留原记录、调用配对和计费，避免并行或跨轮结果累积超窗。
Fit cumulative tool results before each model request while preserving the transcript, call pairing and usage.
Verified with 121 related tests, 40 stress scenarios, Pi HTTP failure injection and independent reviews.
Preserve partial output and token usage in CLI turns and edited UI rounds while leaving ordinary failures and cancellation unchanged. Use OutputLimitError across agent runners so incomplete turns remain visibly unsuccessful but can be reloaded and continued.
Cover real stream handling, SQLite persistence, reload, CLI continuation, and unsuccessful one-shot JSON output.
Preserve repeated same-second WeFlow and EchoTrace messages while consuming existing fallback occurrences one for one across overlapping imports and merged sources.
Keep compression summaries before the in-flight question and let the Node runtime validate and atomically replace only the terminal question-answer round.
Replace the edited user message, old assistant response, and regenerated response in one SQLite transaction. This prevents a failed replacement insert from leaving a partially rewritten conversation.
中文：将用户与助手消息合并为单事务写入，失败时不再留下孤儿消息或误报成功。
English: Persist user and assistant messages in one transaction so failures leave no orphaned turn or false success.
中文：避免同秒消息跨摘要边界，防止已压缩内容重复回放或后续消息丢失。
English: Keep equal-timestamp messages on one boundary side so compressed content is not replayed and later messages remain visible.
Use dependency-independent Chinese tokenization, preserve prompt capacity for the highest relevance match, and handle short symbolic and numeric terms without distorting bounded recall.
中文：同名 LLM 配置此前会覆盖同一 auth profile；现在按稳定 config ID 隔离，并保留旧 profile 恢复。
English: LLM configs with the same name previously overwrote one auth profile; new profiles are scoped by stable config ID while legacy restoration remains compatible.
Remove the proactive-memory preference and its cross-platform runtime plumbing so memory behavior has one clear contract. Normalize away the obsolete setting and keep source type limited to provenance.
Include the normalized time-filter request in structural-inspection fingerprints while retaining the cursor's resolved absolute range, so stable relative pagination works but changed filters are rejected.
Interpret recent-day filters as complete local calendar days and persist the first resolved range in structural-inspection cursors so pagination remains stable across clock and date changes.
Resolve private ranking display names, apply privacy preprocessing consistently across cached recaps and structural labels, and bound model-visible overview payloads to the configured tool budget.
Keep shared-interaction evidence, pagination, and token truncation accurate while bounding anchor inputs, proximity scans, contact discovery, and reply aggregation. Resolve participants across all imported history and isolate damaged database candidates without changing the final analysis tree.
Use the all-history contact snapshot and validate raw session candidates inside the cancellable scan loop. Bound model-visible structural results to the configured token budget while preserving full tool details and continuation metadata.
moesnow
digua
Sanssssssssssssssss
liuyy2
babybitter
KokerZhou
Sanssssssssssssssss