0xzkslr-ai e89b128157

fix(runtime): harden model response recovery at provider boundaries (#5080) ...

* fix(models): preserve DeepSeek thinking tool history

* fix(runtime): harden model response recovery

* fix(runtime): tighten model response recovery

* fix(runtime): protect run-scoped retry state

* fix(runtime): complete model recovery review fixes

* fix(runtime): preserve empty-response diagnostics

* fix(runtime): strip native tool calls on length caps

* docs(middleware): fit recovery guidance within inherited size limit

---------

Co-authored-by: Willem Jiang <willem.jiang@gmail.com>

2026-09-18 07:29:34 +08:00

..

agents

fix(runtime): harden model response recovery at provider boundaries (#5080)

2026-09-18 07:29:34 +08:00

authz

fix(authz): recheck policy before sandbox reuse (#5006)

2026-08-30 15:29:51 +08:00

community

feat(tavily): support configured search domain filters (#5513)

2026-09-18 07:25:49 +08:00

config

feat(projects): Projects MVP Phase 2 — instructions, document shelf, promotion, trash (#5443)

2026-09-16 18:46:18 +08:00

extensions

feat(extensions): expose incremental run evidence reader (#5405)

2026-09-16 21:18:05 +08:00

guardrails

feat(extensions): let an out-of-tree extension observe what the agent did (#4863)

2026-08-23 09:57:12 +08:00

integrations

feat(frontend): move capability management out of Settings (#5468)

2026-09-16 18:39:24 +08:00

mcp

fix(mcp): isolate pooled sessions by owning event loop (#5396)

2026-09-13 17:13:45 +08:00

models

fix(runtime): harden model response recovery at provider boundaries (#5080)

2026-09-18 07:29:34 +08:00

persistence

feat(extensions): expose incremental run evidence reader (#5405)

2026-09-16 21:18:05 +08:00

projects

feat(projects): Projects MVP Phase 2 — instructions, document shelf, promotion, trash (#5443)

2026-09-16 18:46:18 +08:00

reflection

docs: govern agent guidance size (#4799)

2026-08-13 21:49:04 +08:00

runtime

fix(events): preserve DB write-lock generation across deletion (#5462)

2026-09-17 22:06:03 +08:00

sandbox

fix(sandbox): drain previous release during async rebind (#5498)

2026-09-17 22:33:18 +08:00

scheduler

feat(scheduler): add interval schedule type (#5291)

2026-09-10 16:46:01 +08:00

skills

feat(skills): rank deferred discovery by agent intent (#5369)

2026-09-17 07:59:58 +08:00

subagents

fix(subagents): scale max_turns into the graph's super-step budget (#5485)

2026-09-17 09:05:25 +08:00

tools

feat(runtime): persist tool-progress phase transitions (#5214)

2026-09-15 08:45:08 +08:00

tracing

feature(gateway): issue request trace ids unconditionally (#5119)

2026-09-01 16:49:39 +08:00

tui

feat(channels): select custom agents per conversation (#5168)

2026-09-06 16:46:51 +08:00

uploads

feat(projects): Projects MVP Phase 2 — instructions, document shelf, promotion, trash (#5443)

2026-09-16 18:46:18 +08:00

utils

feat(projects): Projects MVP Phase 2 — instructions, document shelf, promotion, trash (#5443)

2026-09-16 18:46:18 +08:00

workspace_changes

fix(workspace-changes): record symlink targets without the verbatim prefix (#5250)

2026-09-08 14:51:57 +08:00

__init__.py

refactor: split backend into harness (deerflow.*) and app (app.*) (#1131)

2026-03-14 22:55:52 +08:00

AGENTS.md

fix(agents): make create_deerflow_agent's subagent limit, summarization and token_budget features take effect (#5488)

2026-09-17 09:35:50 +08:00

checkpoint_patches.py

fix(checkpoint): unwrap Overwrite first writes into empty channels (#4383)

2026-07-26 10:44:50 +08:00

client.py

fix(client): emit text a later node appends to an AI message already sent (#5479)

2026-09-17 08:45:29 +08:00

constants.py

fix(conversation): clarify reference semantics and keep reader pages inline (#5421)

2026-09-14 15:05:30 +08:00

logging_config.py

fix(channels): stream-cap and validate WeChat/WeCom inbound media downloads, fixes #5223 (#5225)

2026-09-16 16:33:47 +08:00

trace_context.py

feature(gateway): issue request trace ids unconditionally (#5119)

2026-09-01 16:49:39 +08:00