Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for policy analysts

Regulatory Comment Docket Agent

A complete response table where every distinct issue in the comments has a counted entry and a drafted answer

Document handlingCreatingGovernment
See how it worksGet access
Regulatory Comment Docket Agent: what goes in, what the agent does and what you get

What it does

A proposed rule can draw hundreds or thousands of comments, many of them copies of the same form letter, and the agency must respond to every distinct issue. This agent collects the comments from the docket and removes exact and near duplicates, keeping a count of how many people signed each version. It groups the rest by issue, notes who the commenters are, such as residents, businesses or associations, and extracts the specific requests. It drafts a response table with each issue, how many raised it and the agency's answer from the draft rule and staff notes. Before finishing, it checks that every distinct issue has a response and that no comment is left out. If one is missed, it adds it and rebuilds the table. The analyst approves the table. Edge case: a long letter with one new point hidden on page nine is not discarded as a duplicate.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueYes, continueApprovedNoNo 1 STARTS WHEN Comment period closes or weekly interim run 2 USES A TOOL Collect all comments from the docket 3 DOES Remove duplicates and count signers per version 4 DOES Group comments by issue and note commenter type 5 CHECKS THE RESULT Does every comment map to at least one issue? If not: create a new issue group or flag the comment forthe analyst. Back to step 3. 6 DOES Draft the response table with counts and proposedanswers 7 USES A TOOL Check each answer against the rule text and staffnotes 8 CHECKS THE RESULT Does every distinct issue have a sourced response? If not: mark the gap and ask the analyst for theagency's position. Back to step 6. 9 YOU APPROVE Analyst approves the response table 10 RESULT Response table filed with the docket
Read the steps as a list
  1. Comment period closes or weekly interim run
  2. Collect all comments from the docket
  3. Remove duplicates and count signers per version
  4. Group comments by issue and note commenter type
  5. Does every comment map to at least one issue?If not: create a new issue group or flag the comment for the analyst. Back to step 3.
  6. Draft the response table with counts and proposed answers
  7. Check each answer against the rule text and staff notes
  8. Does every distinct issue have a sourced response?If not: mark the gap and ask the analyst for the agency's position. Back to step 6.
  9. Analyst approves the response tableThe agent waits here for your OK.
  10. Response table filed with the docket

How it decides

Comments are duplicates when their text matches above a set similarity. Each distinct issue gets its own row, and an issue is answered only from the rule text or staff notes.

  • Treat comments above 90% text similarity as duplicates
  • Read the full text of long letters for hidden new points
  • Give each distinct issue its own row
  • Answer only from the rule or staff notes, never from guesses

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Duplicate similarity threshold (default 90%)
  • Issue categories
  • Commenter types
  • Table format

What keeps you in control

It always asks you first

  • The response table before publication

Hard limits

  • Never publishes the table
  • Does not decide the agency's position

It stops when

  • Done: all issues have responses and the table is approved
  • Stop: docket export is incomplete

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensA zoning docket drew 612 comments. 402 were one form letter, 88 were a second, leaving 122 unique letters. The agent grouped them into 17 issues. In the check, 9 comments matched no issue, so it created two more groups. One letter raised a parking issue on page nine, which became row 18. The analyst approved a table of 19 issues with counts.

More agents for policy analysts

Reporting · Research

Stakeholder Brief Update Agent

Current, documented stakeholder briefs.

8 steps1 check5 jobs
Reporting · Government

Campaign Promise Delivery Tracker Agent

A sourced status table for every promise that the official can publish with confidence

10 steps2 checks4 jobs
Creating · Automation

Amendment Comparison and Impact Agent

A short briefing note on what each amendment changes, with conflicts and knock-on effects listed and sourced

11 steps2 checks3 jobs
Analysis · Quality control

Fiscal Note Review Agent

A memo that shows whether the fiscal note's totals can be reproduced, which assumptions drive them and a range

10 steps2 checks3 jobs