Keep a sheet of new court decisions on your research question
Each week, searches the legal source you name for your research question and records each decision's court, date and outcome in a sheet.
Apps it connects
Authenticate each one once and the workflow is ready to run.
What it does
Each integration step this workflow runs, in order, what it is for, and the conditions and loops that decide when it runs.
- 1Google SheetsRead rowsReads the 'Case notes' tab to collect the addresses already recorded, so no case is written up twice.
- 2SerperSearch webRuns your research question as a search limited to the one legal site you named, returning the decisions to read.
For each
result- 3Jina AIRead urlFetches the text of each decision page found on the site you named, so only that page can be quoted.
If the length of
pageTextis at least 400- 4AnthropicGenerate model responseAsks the AI to copy the case name, citation, court, date, stated outcome and one passage from the page text alone.
If the number of items in
partsis at least 6- 5Google SheetsAppend rowAppends one row per case to the 'Case notes' tab, leaving cells empty for anything the page did not state.
- 5
- 4
- 3
How it works
What happens on each run, in plain English.
Once a week, your standing research question is searched on the one legal website you name, each decision found there is opened, and what the page states is copied into a row of your spreadsheet. It is a research aid, not legal advice, and it never says what a case means or how it applies.
You end up with one row per case in a 'Case notes' tab: case name, citation, court, decision date, the outcome as the page states it, a quoted passage, the source address and a note of what the page did not say. Only pages on the site you named are ever opened. Anything the page does not state is left empty, never guessed, and the quoted passage is checked against the text actually fetched: if those words are not on the page, the quote and outcome are emptied and the row says so. Cases already recorded are skipped, and each run reads a capped number of pages.
Before the first run you need a spreadsheet with a 'Case notes' tab, a research question, and the host name of the one site to read.
You may also be interested in
Turn regulatory updates into tracked tasksReads the official sources you list each day, summarises every new or amended rule, and files a task for each one.
Never miss a filing or renewal deadlineChecks your deadline sheet each morning and emails you as each licence, permit, tax filing or renewal comes due, with a calendar event.
Handle a data erasure request, with approvalFinds every record for one person, shows exactly what it found, and erases it only after a person approves.Fill in what each company sells, and who toReads each company in your sheet from its own web page and records what it sells and who to, with the page it read.Start from a workflow that already works.
Add "Keep a sheet of new court decisions on your research question" to your workspace, connect its apps, and make it yours.
Free plan, no credit card