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 public affairs specialists

Crisis Recovery Sentiment Tracking Agent

The manager sees an evidence-based recovery picture each day, with causes behind any setback.

Crisis Recovery Sentiment Tracking Agent: what goes in, what the agent does and what you get

What it does

After an incident, leaders ask whether things are getting better, and the answer is usually a feeling. This agent measures it every day against a pre-incident baseline: coverage tone, share of voice, search suggestions and the main themes in social posts. It records the numbers and compares them with the baseline and the previous day. If a measure moves the wrong way, it digs into the posts and outlets driving it to find the cause, then checks whether the cause is a single outlet or a wider group. It drafts a weekly recovery report with a suggested next step. The manager approves the report and any new response. Edge case: a spike comes from one viral post, so the agent labels it isolated.

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 Daily tracking run 2 USES A TOOL Collect coverage, share of voice, search suggestionsand social posts 3 DOES Compute each measure against the baseline andyesterday 4 CHECKS THE RESULT Did any measure worsen by 10 percent or more? If not: if data looks incomplete, refetch the last 24hours before moving on. Back to step 2. 5 USES A TOOL Open the posts and outlets driving the change 6 DOES Classify the cause as isolated, a group or a newstory angle 7 CHECKS THE RESULT Is the cause supported by at least 3 examples? If not: pull more samples and reclassify. Back to step5. 8 DOES Draft the daily note and weekly recovery report witha suggested step 9 YOU APPROVE Manager approves the weekly report and any newresponse step 10 RESULT Report filed and baseline comparison updated
Read the steps as a list
  1. Daily tracking run
  2. Collect coverage, share of voice, search suggestions and social posts
  3. Compute each measure against the baseline and yesterday
  4. Did any measure worsen by 10 percent or more?If not: if data looks incomplete, refetch the last 24 hours before moving on. Back to step 2.
  5. Open the posts and outlets driving the change
  6. Classify the cause as isolated, a group or a new story angle
  7. Is the cause supported by at least 3 examples?If not: pull more samples and reclassify. Back to step 5.
  8. Draft the daily note and weekly recovery report with a suggested step
  9. Manager approves the weekly report and any new response stepThe agent waits here for your OK.
  10. Report filed and baseline comparison updated

How it decides

A measure is a setback when it worsens by 10 percent or more from the previous 3-day average. It digs into the top sources behind it.

  • Flag a setback at a 10 percent worsening against the 3-day average.
  • Treat a cause as real with at least 3 examples.
  • Call recovery when all measures sit within 15 percent of baseline for 7 days.
  • Escalate a new angle to the manager the same day.

Make it yours

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

  • Measures tracked
  • Setback threshold (default 10 percent)
  • Recovery definition (default 7 days within 15 percent)
  • Report recipients

What keeps you in control

It always asks you first

  • Manager approves the weekly report and any new response step

Hard limits

  • Report numbers with their source
  • Do not respond publicly on its own

It stops when

  • Done: all measures near baseline for 7 days and manager closes tracking.
  • Stop: monitoring tools down, so the manager is told.

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 happensOn day 9 after the incident, negative tone rose from 22 to 31 percent. The cause check first found only 1 post, so it failed. The agent pulled 40 more samples and found 6 posts from a customer forum about refunds. It classified it as a group issue and suggested a refund FAQ. The manager approved the response and the weekly report.

More agents for public affairs specialists