Choose one of three approaches. The steps for that option appear below.
Choose an option above. On a phone, swipe the tabs left or right to see all options →
Every trigger has an observable condition and source coverage limits; no watch is activated. The topic lead signs off.
Assemble the source set
Draft the ledger
Check rows against originals
Draft the specification
Lead signs off draft
Use: You
Assemble the research question, trusted source list and relevance rules as one source set; record scope, permission limits and source IDs.
Check every file opens and is lawfully accessible or owner-authorised.
Use: Gemini Notebook
Load the whole source set and ask for the evidence ledger in one pass.
What to paste into the tool
Using every supplied source, build a ledger. Evidence ledger rows with source_id, source_date when known, passage_or_cell_location, extracted_value_or_quote, evidence_status, permission_note and unresolved_question; original units and time zones kept. Use only the supplied material. Quote exactly and give the source ID and location for every line. Separate what a source says, what is inference and what is not known yet. Do not send, publish, buy, book, change any account or run code.
Use: You
Check every ledger row against the originals and mark rows that fail.
Compare every extracted passage, date, identifier and numeric field with the original; flag missing or unreadable inputs.
Use: Gemini Notebook
Ask for the watch specification: translate the question into inclusion rules, exclusions and materiality thresholds.
What to paste into the tool
From the checked ledger only, draft the watch specification. Write each trigger as an observable condition and state the coverage limits of each listed source. Prepare the specification only: do not set up, activate or schedule any monitor, and do not imply real-time coverage. Use only the supplied material. Quote exactly and give the source ID and location for every line. Separate what a source says, what is inference and what is not known yet. Do not send, publish, buy, book, change any account or run code.
Use: The topic research lead
Check every trigger condition and its source coverage, record unresolved questions, and sign off only the draft watch specification. No watch is activated.
Every trigger has an observable condition and source coverage limits; no watch is activated. Only the checked ledger reaches it.
Fill the ledger by hand
Second reader checks
Draft the specification
Trace lines, sign off draft
Use: You
Read the originals and fill the evidence ledger by hand in a spreadsheet; no AI tool sees the raw sources.
Compare every extracted passage, date, identifier and numeric field with the original; flag missing or unreadable inputs.
Use: The second reader
Check the ledger against the originals and return disputed rows; confirm each quote, date and number matches the source; list anything unreadable.
Use: ChatGPT
Give only the checked ledger; ask for the watch specification citing ledger row IDs.
What to paste into the tool
You have only the ledger below, not the sources. Draft the watch specification citing ledger row IDs. Write each trigger as an observable condition and state the coverage limits of each listed source. Prepare the specification only: do not set up, activate or schedule any monitor, and do not imply real-time coverage. For anything the ledger does not support, write a question and do not fill it in. Use only the supplied material. Quote exactly and give the source ID and location for every line. Separate what a source says, what is inference and what is not known yet. Do not send, publish, buy, book, change any account or run code.
Use: The topic research lead
Trace each line back to the original; record unresolved questions and sign off only the draft watch specification. No watch is activated.
Every trigger has an observable condition and source coverage limits; no watch is activated. You write the specification.
Split pack per source
Source card per chat
Combine cards by hand
Draft specification by hand
Challenge the draft
Lead signs off draft
Use: You
Split the pack into one file per trusted source and register each file.
Check no file lost pages or tables in the split and that permission limits carry over to each file.
Use: Claude
For each trusted source, open a fresh chat with only that file and ask for a trusted source card.
What to paste into the tool
This chat holds one trusted source only. List exact passages relevant to the research question with source ID and location, and mark each as stated, not stated or ambiguous. Use only the supplied material. Quote exactly and give the source ID and location for every line. Separate what a source says, what is inference and what is not known yet. Do not send, publish, buy, book, change any account or run code.
Use: You
Combine the trusted source cards into the ledger by hand and note conflicts between sources.
Compare every extracted passage, date, identifier and numeric field with the original; flag missing or unreadable inputs.
Use: The topic research lead
Draft the watch specification by hand from the ledger: translate the question into inclusion rules, exclusions and materiality thresholds.
Use: Claude
Challenge the human draft: list lines with no ledger support and requirements or criteria it does not cover.
What to paste into the tool
Review the draft against the ledger. List every line without a ledger row, every missing item, and anything stated more strongly than the ledger allows. Do not rewrite the draft. Write each trigger as an observable condition and state the coverage limits of each listed source. Prepare the specification only: do not set up, activate or schedule any monitor, and do not imply real-time coverage. Use only the supplied material. Quote exactly and give the source ID and location for every line. Separate what a source says, what is inference and what is not known yet. Do not send, publish, buy, book, change any account or run code.
Use: The topic research lead
Act on the challenge list, then record unresolved questions and sign off only the draft watch specification. No watch is activated.
Best when: One assistant drafts both from one source set
What it does: Gemini Notebook drafts ledger and specification; people check
Strong at: A single trail from source to draft
Limit: Completeness rests on the analyst's row check
Free plan: Yes, limited (50 sources per notebook)
Best when: Raw sources stay with people
What it does: ChatGPT works from the checked ledger only
Strong at: Each draft claim traces to a ledger row
Limit: Anything missed in the ledger is unrecoverable
Free plan: Yes
Best when: Distinct sources, human authorship wanted
What it does: Claude reads per source and challenges; people write
Strong at: Authorship and judgement stay with the reviewer
Limit: 20 files per chat, 1000 pages per PDF; matters for large files
Free plan: Yes, limited (session usage cap resets every 5 hours)
Before you paste into any AI tool, remove any confidential information. Put it back at the end.
Deciding between ChatGPT and Claude for this? See the comparison.
Use only material you may lawfully use or have the owner's permission to use, and keep the source IDs and where each passage or cell came from. Treat the output as a research suggestion for your review.
This page prepares a watch or analyses snapshots you supply. It does not start an ongoing monitor or mean real-time coverage.
Do not let the tool send, publish, buy, book, change an account, start monitoring or run live code on its own. You keep every decision and any later approval.
Give every trigger an observable condition; note the limits of source coverage. Settle the evidence trail and open items, and approve any later monitoring separately, before the brief is complete.
Check the result against your original inputs. Approve it before sharing or acting.