scansci-pdf

Files (1)

• SKILL.md

SKILL.md (first lines)

# ScanSci PDF

Use the bundled `scansci-pdf` MCP server for paper discovery, metadata, citations,
download queues, access diagnostics, and result reporting. The repository's full
workflow reference is `../../skill/SKILL.md`; read it when the task needs detailed
tool routing or configuration guidance.


## ⚠️ 修改反爬/嗅探/镜像相关代码前必读

先读 [`docs/PLAYBOOK.md`](../../docs/PLAYBOOK.md)：镜像健康唯一存储（domain_db wall_state 表）、
结构性失败签名表、节奏参数、已知死镜像清单、以及六条铁律（含"造轮子前先 grep 家底"——
镜像健康存储曾被重复造过三代）。
## Route by intent

- Search or identify papers: `scansci_pdf_search`, `scansci_pdf_sciverse`
  (`search`/`content`/`meta_search`/`relations`/`catalog`/`resource`; requires
  `sciverse_api_key`), `scansci_pdf_prepare_queue(action="verify")` for
  identifiers, and `scansci_pdf_parse_list` for list files.
- Configure Sciverse (optional semantic evidence/metadata): `scansci_pdf_sciverse_setup`
  with `token` and `test=true`; without it, search still works and returns a `config_hint`.
- Resolve open-access locations: `scansci_pdf_prepare_queue(action="resolve_oa")`.
- Download one paper: `scansci_pdf_download` with an explicit strategy when the
  user has a source preference.
- Process a list: `scansci_pdf_parse_list` or `scansci_pdf_batch_download`.
- Export citations: `scansci_pdf_citation`; push to Zotero via `scansci_pdf_zotero_push`.
- Diagnose setup or access: `scansci_pdf_diagnostics` (check=health|network|sources|setup).

## Operating boundaries

- Preserve the user's requested source strategy and report the actual source in
  every download result.
- Prefer open-access, publisher API, and institution-authorized routes when no
  source preference is given.
- Institutional access turns on two things users routinely get wrong, so state
  them rather than let the attempt fail: the login has to happen in the browser
  the tool opens (`scansci_pdf_login`) because cookies are captured from that
  context — a headless window, or the user's own browser, yields no session —
  and the **egress IP is what the publisher sees**, so a system-wide proxy moves
  it off campus and silently voids IP-based entitlement. When a publisher
  refuses a PDF that should be entitled, read the IP on the publisher's error
  page before concluding the institution lacks a subscription.
- Ask before opening an interactive login, importing cookies, or changing access
  configuration.
- Treat API keys, cookies, proxy credentials, and institution details as secrets;
  never repeat them in the response.
- Gray-source and anti-bot routes require the user's explicit choice and must be
  used only where the user has the right to do so and applicable rules permit it.

From Rimagination/scansci-pdf, MIT.

More from Rimagination/scansci-pdf