Laid off? The 30-day plan to land a new job with AI

Goal for the first two days: 3 free AI certifications on your profile. They prove AI fluency before the first interview even happens.

Anthropic Academy (best for AI fluency):

• AI Fluency: Framework & Foundations (4 to 6 hours, self-paced)
• Claude 101 (2 to 3 hours, 13 lessons)
• Introduction to Claude Cowork (2 hours)

Microsoft + LinkedIn (most recruiter recognition):

• Career Essentials in Generative AI (professional certificate)
• AI for Managers (4h 34m)

The fastest ones if you want volume: the Salesforce Trailhead Generative AI Basics badge takes 20 minutes, and the self-paced AWS Generative AI for Executives takes about 50 minutes.

Recruiters do not read your profile. They find it (or not) through keyword searches and skills filters. Give Claude your profile with 3 dream job descriptions, and let it optimize every section.

The complete prompt to paste into Claude:

You are my LinkedIn rewriter. I just got laid off. My job is to look hireable for
recruiters searching for AI-skilled candidates in my field within 14 days.

INPUTS (I'll paste these in order):
1. My current LinkedIn profile (headline + About + Experience + Skills) as plain text
2. Three job descriptions for roles I'd take in a heartbeat
3. The 3 most senior roles I've held with actual measurable outcomes (numbers, not duties)

Run this exact protocol:

STEP 1. KEYWORD MAPPING. Pull every recurring word and phrase from the 3 JDs that
appears in 2+ of them. Group into: must-have skills, nice-to-have skills, soft skills,
AI-specific terms. Show me the keyword cluster before you rewrite anything.

STEP 2. HEADLINE REWRITE. Write 3 versions of my LinkedIn headline using the keyword
cluster. Each version must:
- Be under 220 characters
- Start with my actual role + outcome, not a label
- Use the 3-4 keywords recruiters are most likely to search
- Avoid corporate words ("results-driven," "passionate about," "innovative")
- Read like a real person wrote it

STEP 3. ABOUT REWRITE. Write the About section so the first 3 lines (above the "see more"
fold) do all the work:
- Line 1: who I am + the outcome I create
- Line 2: the proof (a specific metric or accomplishment)
- Line 3: what I'm looking for next
Then elaborate. Total under 1,500 characters.

STEP 4. SKILLS SECTION. LinkedIn lets you list 100 skills. Most people list 10. Generate
the full 100, ranked by recruiter-search frequency for my field. Pull from the keyword
cluster + broader AI/role vocabulary. Recruiters filter by Skills more than they search
by hashtag. This is the most important section.

STEP 5. EXPERIENCE BULLETS. Rewrite my top 3 roles using (Action + Outcome + Metric).
Strip every "responsible for" and "duties included." Lead with verbs. Lead with results.
Cap each role at 5 bullets.

OUTPUT FORMAT: Show me each step in order. Don't compile the full profile at the end.
I want to review and edit each section as we go.

Constraints:
- Don't invent metrics I didn't give you.
- Don't add skills I don't have.
- If a JD requires a skill I clearly don't have, flag it. Don't fake it.

Between days 5 and 10, you build ONE public proof of AI fluency: a tool that solves a real problem from your last job. Not a generic demo.

The complete prompt to paste into Claude:

You are my AI portfolio strategist. I just got laid off and I need to ship one visible,
tagged, LinkedIn-postable AI project in the next 5 days. The goal: recruiters who land
on my profile see proof of AI fluency, not just claims.

Here's my situation:
- My previous role: [PASTE TITLE + 2-LINE DESCRIPTION]
- The painful problem from that job I'd build a tool to solve: [PASTE. Be specific.
  "The Monday report that took me 4 hours every week" or "the inbox triage I'd kill for"]
- The 3 roles I'm targeting: [PASTE OR LIST]

Run this protocol:

STEP 1. PICK THE FORMAT. Based on my role and target roles, recommend ONE:
- A Custom Claude Skill (best for ops, EA, marketing, analyst, PM, CS)
- A working Claude Artifact: dashboard, calculator, tracker, generator (best for finance,
  data, product, consulting, founder roles)
- A scheduled automation that runs a workflow in the background (best for senior IC, eng
  lead, director)
Pick one. Tell me why. Don't list all three and let me decide.

STEP 2. BUILD SPEC. For the format you picked, give me:
- The exact problem in 1 sentence
- The user (who would use this)
- The 3-5 inputs the project needs
- The 3-5 outputs the project produces
- The mechanic (which Claude features it uses)
- The "would a real person use this?" test. Answer honestly.

STEP 3. BUILD IT WITH ME. Generate the actual Claude Skill prompt, Artifact prompt, or
scheduled task setup. Make it production-quality. I should be able to copy your output,
paste it into Claude, and get a working tool in under 30 minutes.

STEP 4. THE POST. Draft the LinkedIn post I'll publish when I share it:
- Hook: 1 line about the problem
- Body: 3-5 lines about what I built and what it does (specific, with numbers if possible)
- Visual: tell me what to screenshot (a real output, not a stock graphic)
- Hashtags: #AIProjects #GenerativeAI #BuildInPublic + one role-specific tag
- CTA: invite people to DM if they want the build
- Length: under 1,300 characters (LinkedIn fold limit)

Constraints:
- The project must solve a real problem from my last job. Not a generic demo.
- Imperfect public beats perfect private.
- If I list a problem you can't solve with Claude alone, say so and adjust scope.

From day 10 to day 30, the rhythm is simple: 5 applications a day, each one tailored. Without AI, a good application takes 60 minutes. With this prompt, 10.

The complete prompt to paste into Claude:

You are my application factory. I'm applying to 5+ jobs a day and each one needs a
tailored resume + cover letter. The bottleneck is time, not effort. Your job: cut each
application from 60 minutes to 10.

ONE-TIME SETUP. I'll paste once:
- My master resume (the long version with every role and every bullet)
- My base cover letter template (the 200-word version with my pitch)
- My about-me.md file (so you know my voice)
- My non-negotiables (salary minimum, remote vs. hybrid, locations, role types I won't take)

After that, every application starts with me pasting the JD. Run this protocol:

STEP 1. JOB FIT SCORE. Read the JD and tell me, in one paragraph:
- The 3 keywords this role is searching for
- My fit score (1-10) based on my master resume
- The 2-3 things I'm genuinely strong on for this role
- The 1-2 places I'm a stretch
- Whether I should apply or skip. If skip, why. If skip, stop here.

STEP 2. TAILORED RESUME. Take my master resume and:
- Reorder so the most relevant role is on top
- Rewrite 2-3 bullets per role to match the JD's vocabulary (don't invent, remap what's there)
- Trim everything not earning its place
- Hit the JD's keywords in the Skills section
- Output a 1-page version

STEP 3. TAILORED COVER LETTER. Rewrite my base template:
- Open with a specific reference to the company or the role (not "I'm excited to apply")
- Address the 2-3 strengths from Step 1
- Briefly handle the stretch (don't dwell)
- Close with a clear next step
- Length: 200-300 words
- Voice: match my about-me.md exactly.

STEP 4. APPLICATION NOTE. Generate a 1-line note I can paste into any "anything else?"
box. Specific. Memorable. Not a re-pitch.

STEP 5. OUTPUT FORMAT. Give me all three pieces in one response: fit score + verdict,
resume in plain text, cover letter in plain text, application note. Ready to paste into
the form.

Constraints:
- Don't fabricate experience.
- If the JD is impossible to read or vague, flag it.
- If the company has a known controversy I should know about, mention it briefly.

Before every interview, you give Claude the job description, your tailored resume and the interviewer's LinkedIn. It predicts the questions, writes your answers, then makes you rehearse out loud.

The complete prompt to paste into Claude:

You are my interview prep coach. I have an interview for [ROLE] at [COMPANY]
on [DATE]. Your job: generate the 30 most likely questions, the strongest answer
for each based on my real experience, then mock-interview me by voice.

I'll paste:
- The job description
- My tailored resume (the one we built for this role)
- The LinkedIn profile of the interviewer (or interviewers, if there's a panel)
- The 3 hardest things I might be asked about my background (gaps, transitions, layoff,
  career change)

Run this protocol:

STEP 1. 30 QUESTIONS. Generate 30 likely interview questions, ranked by probability:
- 10 behavioral (STAR method territory)
- 10 role-specific (technical or domain)
- 5 about my background (the gaps + the layoff)
- 5 culture / motivation ("why us, why now, why this role")

STEP 2. MY STRONGEST ANSWER. For each, write the answer I should give in 90 seconds or
less, using STAR (Situation, Task, Action, Result) where it fits. Use my actual
experience from the resume. Don't make up stories. If I don't have a real example for a
question, say so and suggest how to handle the gap honestly.

STEP 3. THE LAYOFF QUESTION. Specifically, write me 3 versions of the answer to "Why did
you leave your last role?". Short, direct, no apology, no oversharing. I'll pick the
one that fits.

STEP 4. THE WILDCARD. Tell me the one question this interviewer is statistically most
likely to ask based on their LinkedIn (their background, their team, their recent posts).
Prep me specifically.

STEP 5. VOICE MOCK. Once we've gone through the questions, switch to voice mode. Ask me
10 random questions from the list, one at a time. After each answer, give me one specific
note. Not "great answer," real feedback. ("You hedged in the middle. Cut the 'I think.'"
"You ran 2 minutes. Cut to 90 seconds.")

Constraints:
- Make the questions sound like real interviewers.
- Don't let me get away with vague answers. Push for specifics.
- Don't pep-talk me. Coach me.