科研 Skill
法律研究插件初始化配置
setup
One-time, conversational setup for the legal-research plugin: saves and tests a CourtListener token and checks Zotero credential availability. Use to set up this plugin’s CourtListener access.
SKILL.md
Setup
The user is probably a law professor, not a programmer. Do the work for them, explain
each step in one plain sentence, and never ask them to open a file, the Terminal, or a
text editor unless a step says so explicitly. Everything below is done with
../setup/setup.py (adjust the path to wherever this
plugin's files actually live if the current agent does not expose
${CLAUDE_PLUGIN_ROOT}), which backs up any file it rewrites.
This script is agent-neutral: it never reads or writes a host agent's own settings
file. It does not touch ~/.claude/settings.json, ~/.claude.json, or any
marketplace/auto-update registry — those are specific to one Claude Code install, and
either do not exist or mean something different in another agent. Two things the
original upstream version of this skill used to configure automatically are not
handled here and are instead the user's own job, in whatever menu the current agent
exposes for it:
- Web access permissions — if the current agent asks for approval before
WebFetch/WebSearchcalls, a long research run can be interrupted by that prompt. Tell the user (once, briefly) to check whether their agent offers a way to pre-approve those tools for a session (in Claude Code:permissions.allowin Settings, or picking "Always" the first time it asks) so a long run isn't interrupted partway through. - Multi-agent / workflow features — if
cite-check,proofread, orlegal-researchdispatch multiple subagents or a bundled.workflow.jsscript and the current agent has a setting that gates that (Claude Code's Dynamic workflows), tell the user to check their agent's own settings for it; this script does not flip that switch.
Run the steps below in order. Skip anything the status report shows is already done, and say so in a few words, such as "CourtListener, already working." Keep the whole exchange short.
1. Take stock
python3 "../setup/setup.py" status
- If
python3is missing or the command fails to start: on a Mac the first use ofpython3opens a dialog offering to install the command line developer tools. Tell the user to click Install, wait for it to finish (a few minutes), then ask again. On Windows, tell them to install Python from python.org (tick the option that adds python to PATH), restart their terminal/app, and run setup again. - If
python.okis false (older than 3.11): same advice, with the current python.org installer. - Do not paste the raw JSON at the user. Summarize it as a short checklist with a tick or cross per line: Python, CourtListener, Zotero (marked optional), git.
2. CourtListener token (required for case checking)
If courtlistener.works is true, say so and move on.
Otherwise explain, in two sentences, that CourtListener is the free, nonprofit case-law database the skills use to confirm that a cited case exists and says what the draft claims, and that it needs a free account. Then give the user these exact steps and wait for their reply:
- Open https://www.courtlistener.com/register/ and create a free account (or sign in).
- Open https://www.courtlistener.com/profile/api-token/ and copy the long string of letters and numbers shown there.
- Paste it here, into this chat, and press Return.
When they paste it, run:
python3 "../setup/setup.py" courtlistener "<token>"
The script tests the token with a live lookup of Brown v. Board of Education before
saving it to ~/.config/courtlistener/token, readable only by the user. Exit 0 means it
works, and it works immediately, with no restart. On HTTP 401 the token was mistyped or
truncated: ask them to copy it again. On a network error, say CourtListener could not be
reached and offer to retry.
Tell the user, once, that the token now lives on their own computer, that the pasted copy also sits in this conversation's history, and that if that ever worries them they can generate a new token on the same CourtListener page and run setup again.
3. Zotero (optional; many users do not have it, and nothing else depends on it)
Say first that this step is optional and that every skill except build-library works without Zotero, checking articles against the open web instead. Then ask whether they keep their sources in Zotero, and say that if so, cite-check and find-cite can search it before the web once it is configured — and that they can answer skip now and come back later. Treat anything but a clear yes as skip, and never present a missing Zotero as a problem in the final checklist; show it as not connected (optional).
Run python3 "../setup/setup.py" zotero-check (read-only) to
see whether ZOTERO_API_KEY / ZOTERO_LIBRARY_ID / ZOTERO_LIBRARY_TYPE are already
set as environment variables. If configured is already true, say so and move on.
Otherwise, this script cannot set those variables for the user — doing so would mean writing to a file specific to one agent (e.g. Claude Code's MCP server config), which this plugin no longer assumes. Instead:
- Have the user get their numeric Zotero
userIDand a private API key from https://www.zotero.org/settings/keys (for a shared group library, the group's number from itszotero.org/groups/NNNNURL instead of the personal userID, plusZOTERO_LIBRARY_TYPE=group). - Tell them to set
ZOTERO_API_KEY,ZOTERO_LIBRARY_ID, and (for a group library)ZOTERO_LIBRARY_TYPE=groupas environment variables in however the current agent exposes that — a shell profile, a.envfile the agent loads, or that agent's own settings UI for environment variables or MCP servers. Point them at that agent's own documentation rather than guessing a path; this plugin's own scripts (build-library/attach_pdf.py,cite-check's Zotero lookup) read those same three variable names directly, so no further configuration inside this plugin is needed once they are set. - Tell them the variables take effect the next time a script in this plugin runs that needs them — usually immediately in a fresh terminal/session, or after the current agent restarts if it caches its environment at startup.
4. Finish
Re-run status and show the final checklist (Python, CourtListener, Zotero). Then, in
two sentences: they are ready; to try it, open a folder that holds a short draft and ask
to cite-check it. If this run set up Zotero, remind them it takes effect next time a
script needs it, per step 3.
Do not install anything else, do not touch files other than the CourtListener token file, and do not read or echo the token or key back after saving them.
示例提示词
帮我配置好CourtListener令牌并完成法律研究插件的初始设置。
所属插件
安装该插件会一并获得同领域的 13 个 Skill / MCP。
同插件的其他科研 Skill
- Zotero文献库构建Builds or extends a Zotero collection for a project: finds sources, adds metadata via the Zotero Web API, and runs an anti-hallucination check. Use to populate a Zotero library with verified sources.
- 裁判文书实证研究Guides empirical research on a sample of Chinese court judgments: sampling frame, coding scheme, inter-coder reliability, and descriptive statistics on rulings. Use for a quantitative case-law study.
- 最高法院案件材料整理Gathers the full document set for a US Supreme Court case and loads it onto the user’s reMarkable in a per-case folder. Use when prepping a SCOTUS case for a podcast, class, or article.
- 文稿引注核查Verifies every citation/quotation in a legal draft against CourtListener, official reporters, and Zotero/the open web, returning a graded Bluebook footnote report. Use to cite-check a US law draft.
- 代码专利特征扫描Scans a codebase for distinctive, potentially patentable design patterns with structured scoring and evidence. Use before a patent consultation, not as legal advice.
- 代码专利检索验证Turns code-scan findings into patent-database search queries for researching existing implementations before consulting an attorney. Not legal advice.
- 快速引注查找Finds 2-4 verified US case-law or scholarly citations with a pincite for a single proposition, footnote-ready with Bluebook form. Use for a quick cite-finding request, not a full memo.
- 中国法律全领域检索China-focused systematic legal research across statutes, regulations, judicial interpretations, and guiding cases, tailored by user role. Use to judge legality or find a legal basis/damages standard.
- 深度法律研究备忘录Fans out multiple agents across free US legal sources, verifies every authority against CourtListener/Justia/Google Scholar, and returns a polished memo. Use for US case-law research.
- 专利构思特征分析Analyzes a described concept to surface what makes it distinctive, with structured scoring for a patent consultation. Not legal advice.
- 专利构思检索验证Turns a concept analysis into patent-database search queries for researching the prior-art landscape before consulting an attorney. Not legal advice.
- 文稿校对润色Proofreads a legal draft’s prose for typos, grammar, and internal consistency (hyphenation, defined terms, cross-references). Use for a proofing pass; not a cite-checker or editor.
常见问题
法律研究插件初始化配置能做什么?
One-time, conversational setup for the legal-research plugin: saves and tests a CourtListener token and checks Zotero credential availability. Use to set up this plugin’s CourtListener access。
如何安装法律研究插件初始化配置?
在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「法律研究插件初始化配置」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。
法律研究插件初始化配置需要哪些权限?
该科研 Skill没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。
法律研究插件初始化配置经过官方审核吗?
是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。