[F170] LangGraph 驱动的研究节点持久化与真实 Web Search
- Dominant language
- TypeScript
- Stars
- 0
- Forks
- 0
- Avg merge
- 1h 7m
- Merged PRs (30d)
- 969
Description
## 用户目标
用户要求将引导式 Deep Research 的每一步拆成 LangGraph 节点;每一步数据属于 Graph State;每次前端节点写调用提交当前完整 nodeState;前后端持久化并可恢复。F170 继续负责按确认大纲执行真实 Web Search、章节进度、来源与局部失败恢复。
## 设计
设计规格将落在 docs/superpowers/specs/2026-08-15-guided-research-langgraph-persistence-design.md。Graph 基础、Brief、Directions、Outline、Research、Report 按独立垂直切片串行交付,F170 不偷并 F171 报告范围。
## 队列外原因
用户在当前会话明确要求先拉最新代码并继续完成 Research LangGraph 持久化,因此登记并公开该工作;正式 claim 前仍需先解决 F180 已合入但 feature_list 仍为 in_progress 的控制平面漂移,并完成契约束修订签核。
## 分支
worker/coord-deep-research-01-f170-langgraph-search
## 验收基线
沿用 feature_list.json 中 F170 的 user_visible_behavior 与 verification;设计确认后补充 LangGraph checkpoint、完整 nodeState、幂等重放、跨组织和单页不刷新反证。
Contributor guide
No contributing guide indexed for this repository
Research direction
Start with docs/superpowers/specs/2026-08-15-guided-research-langgraph-persistence-design.md and the F170 entry in feature_list.json. Resolve the F180 control-plane drift and complete contract review before implementing the Research slice. Done means the existing F170 user-visible behavior and verification pass, with checkpointing, complete nodeState persistence, idempotent replay, cross-organization coverage, and no-refresh evidence.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- typescript
- Domain
- backend-api-design, databases
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 30/100