Skills on AI

Reputation Radar

Download raw SKILL.md

Use when asked to set up a recurring check of how one appears, online (searching one's name, name alerts, old profiles, tagged posts) and in practice (asking trusted colleagues what they would say), or when preparing for a job search or public role. Distinct from [[linkedin-refresh]] (updating one profile), [[getting-feedback]] (receiving feedback on work), and [[identity-protection]] (guarding against fraud and theft).

Reputation Radar is a recurring check of how a person appears to others. It has two halves: what can be found online, and what trusted people would say about them in practice. The name is descriptive; it is not a named method. The goal is no surprises: the user finds problems before an employer, client, or stranger does, and knows what story the evidence tells.

Input required

  • The user's names: full name, common variants, former names, and handles they have used.
  • Context: job search, public role, business, or general upkeep.
  • Profiles and accounts they know of, past and present.
  • Trusted people who could give an honest view.
  • Comfort level with public presence, and any safety concerns.

Questions to ask first

Ask in one batch of at most five questions, skipping any already answered:

  1. Why are you doing this now: job search, public role, or upkeep?
  2. What names, spellings, and handles should be searched?
  3. Which old accounts or profiles might still exist?
  4. Is there anything you already worry might appear?
  5. Who are two or three people who would give you honest feedback?

Method

  1. Search the name. Search each name variant in a few search engines, plus image and news results, in a private browsing window so results are less personalised. Add city, employer, or school to separate the user from namesakes. Note what appears on the first pages.
  2. Record findings. List each result as accurate and helpful, accurate but unhelpful, outdated, inaccurate, or not you.
  3. Set name alerts. Use free alert tools from search engines or similar services, by name variant, if the user wants ongoing notice. Describe the capability; the user picks a product.
  4. Review old profiles. Visit old social, forum, and portfolio pages. Delete, hide, or update what no longer represents the user. Check privacy settings and audiences.
  5. Check tagged and shared posts. Review photos and posts others tagged the user in, and the tag-approval settings in each service.
  6. Tidy and strengthen. Fill gaps with accurate material: a current profile (see LinkedIn Refresh), a short personal page, or posts about real work.
  7. Ask people. Ask two or three trusted colleagues or friends: "What three words would you use for me? What do I do well? What should I change?" Record answers using Getting Feedback.
  8. Compare the two views. Note where online presence and spoken reputation differ, and choose what to change.
  9. Respond calmly to inaccuracies. If something is wrong: wait a day; save evidence such as screenshots and URLs; contact the site or author politely with the specific correction; use the platform's report or correction process; and avoid public arguments.
  10. Schedule the next check. Put a recurring reminder in the calendar, such as twice a year or before a job search.

When to get professional help

If the content involves defamation, harassment, threats, leaked private data, or impersonation, stop and consult a qualified lawyer or the relevant authority. Legal options vary by country; check the current official source. If the user feels unsafe, advise contacting local emergency or support services.

Output format

  • Search log: query, result, category, action.
  • Cleanup list with owners and dates.
  • Alert setup steps.
  • Feedback questions and a response log.
  • Response plan for inaccurate content.
  • Next check date.

Ground rules

  • Do not ask for passwords or recovery codes; the user logs in and edits their own accounts.
  • Do not search for or compile information about anyone else.
  • Do not advise deleting evidence of wrongdoing or lying about the past; focus on accuracy and honest presentation.
  • Do not give legal advice; escalate to a professional.
  • Stay calm and factual; do not encourage retaliation.

Self-check before answering

  • Did I separate accurate, outdated, inaccurate, and not-you results?
  • Did both online and in-person views feature?
  • Is the response to inaccuracies calm, documented, and stepwise?
  • Did I point to a professional for legal or safety matters?

Common pitfalls

  • Searching only once, then never again.
  • Missing namesakes or old handles.
  • Arguing publicly and amplifying the problem.
  • Deleting everything, leaving no accurate presence.
  • Ignoring the human side, what colleagues actually think.
  • Over-reacting to a single minor result.

Example prompt

I'm about to start job hunting. Run a reputation check on me: help me search my name and old usernames, set up alerts, review old profiles, and draft questions to ask three former colleagues. If I find something inaccurate, tell me how to respond calmly.

Learn more

View reputation-radar/SKILL.md on GitHub