light-crawl

1# light-crawl
2 
3## 项目目录与续作
4 
5需要落盘或继续已有项目时，先读 [项目目录约定](references/project-layout.md)，按其中的本阶段位置、版本与关联规则保存；用户指定路径和旧项目优先。包内资源相对本 skill，研究产物相对真实项目，二者不要混用。只做局部讨论时无需初始化目录。
6 
7A routing-first crawler skill. The agent's job is to **compile a target into a
8deterministic scraper**, not to click through every page in a chat loop.
9 
10For research, first identify the caller's actual data gap and prefer an existing
11dataset, lawful download or documented API when sufficient. Read
12[research-data.md](references/research-data.md) when collecting research inputs
13or handing data to Literature, Design, Experiment or Analysis. Acquisition is
14an on-demand service in the research network, not a mandatory stage or evidence
15that a hypothesis is true.
16 
17## 0. Non-negotiables
18 
191. **Scope before traffic.** Read `scope.md`. The user's request for identified
20   public data or an existing authorization can establish scope; record its
21   actual basis without asking again. Clarify a material access/scope gap before
22   traffic. `--authorized` records that decision; it does not prove legal rights.
232. **Cheapest lane first.** HTTP before browser, API before DOM, one page before
24   a crawl.
253. **One browser owner per profile.** Never let two tools launch the same Chrome
26   profile. Claim it first:
27   `light-crawl session acquire <profile> --owner <case-id>`, release it when
28   done, and export state/HAR instead of sharing a profile between tools.
29…