Contract / Policy Reviewer Bot (Field Notes Role Template)

Receive contract or policy → bot runs first-pass review flagging unusual clauses → outputs severity-rated issue list for human expert review

Who

Operations, events, or finance teams who need a fast first-pass screen of venue contracts or policies, producing a checklist for human expert review

You get

A severity-rated flag list (HIGH/MED/LOW) covering unusual clauses, missing protections, market-rate outliers, and questions for the counterparty; explicitly labeled as non-legal-advice first pass requiring human expert finalization

You need

Steps

1. Create bot and paste role description

   In your chat platform, create a new bot, locate the system prompt or role description field, and paste the source prompt below in full. Replace {NAME} with the bot name (e.g., Contract Reviewer) and {HUMAN REVIEWER} with the actual person or title responsible for final review. Do not alter the rest of the instruction text.

   Setup guidance · Source text。The bot's system/role description field; use as the role setting after replacing placeholders

   Copy prompt

   You are {NAME}. When a contract or policy arrives, read it fully and
   produce a first-pass review: clauses that are unusual or one-sided,
   protections we'd expect that are missing, numbers that look off
   versus market, and the questions to ask. Rate each flag {HIGH / MED /
   LOW}. State clearly that this is a first pass for {HUMAN REVIEWER} —
   do not present it as legal advice.

2. Set boundaries and approval gates (keep source recommendation)

   Per the source recommendation, keep the default 'Needs approval for' setting: the bot may read and flag without approval, but any action (replying to the counterparty, modifying terms, signing) requires human sign-off. The source explicitly warns: do not relax this gate until the bot has earned trust.

3. Send contract or policy document to trigger review

   Send the full text of the contract or policy to the bot as a message. The source does not specify a required file format; paste plain text or upload a text-extractable PDF. The bot should then run the first-pass review automatically.

4. Receive severity-rated flag list and conduct human review

   The bot's output should include: a severity rating (HIGH/MED/LOW) per flag, specific clause references, an explanation of the anomaly, and suggested questions to ask the counterparty. The source explicitly states this is a 'first pass' and still requires human expert review for nuance and market rates; it must not be treated as final legal advice or signing authority.

Paste the following role description into your bot's system/role field, replace the placeholders, and save:

You are {NAME}. When a contract or policy arrives, read it fully and
produce a first-pass review: clauses that are unusual or one-sided,
protections we'd expect that are missing, numbers that look off
versus market, and the questions to ask. Rate each flag {HIGH / MED /
LOW}. State clearly that this is a first pass for {HUMAN REVIEWER} —
do not present it as legal advice.

Then send the contract or policy text directly to the bot to trigger the review.