cover-letter

name

cover-letter

description

Write deeply personal, human-sounding cover letters for job applications. Use when the user provides a job description and asks for a cover letter, says "write a cover letter for this", "cover letter for [company]", or when a high-value job application needs a cover letter. Generates a tailored cover letter that opens with a real personal experience connecting Harsha to the company, then maps skills to JD requirements. Output is plain text saved to cover_letters/ folder. Integrates humanizer anti-AI patterns so output never reads like AI wrote it.

Cover Letter Generator

Write a personal, human cover letter for Harsha Vardhan Yellela tailored to a specific job description.

Required Reading (Do All Before Writing)

1. Read references/personal-stories.md (bundled with this skill) to find a matching personal experience
2. Read /home/har5ha/Desktop/resume/references/personal_details.md for skills, experience, education
3. Read /home/har5ha/Desktop/resume/references/projects.md for project details and proof points
4. Read the JD the user provided

Harsha's Personal Context (Use This to Generate Story Hooks)

Use this background to craft a genuine, believable personal connection to any company. Pick the angle that naturally fits the company's domain. Don't force it. Don't reuse the same angle twice in a row.

Childhood & Growing Up (Hyderabad, India):

• Born 2001, grew up in Hyderabad, Telangana
• Math kid — won Gold Medal in Indian National Mathematical Olympiad (INMO) at age 11 (2012). Use for: analytics, fintech, quant, math-heavy companies
• Designed a space settlement called "ATOLL" for NSS/NASA Space Settlement Design Competition at age 14 (2015). Use for: aerospace, space tech, NASA, simulation companies
• Grew up around cricket, Bollywood, and gaming on old CRT TVs with cousins. Indian middle-class childhood with computers arriving late but curiosity arriving early
• First real coding was Arduino/ESP32 tinkering — built a home automation system (DOMUS) controlling lights and fans via voice assistants and a Flutter app. Use for: IoT, smart home, hardware, embedded systems companies

College & Early Career (Hyderabad):

• B.Tech in CS from Geethanjali College (2018-2022)
• Built a COVID-19 tracker app during lockdown using Flutter — first time building something people actually used. Use for: healthtech, public health, mobile app companies
• EPAM Java apprenticeship during college — first taste of professional software engineering, learned what production code really means
• Infor (2022-2023): First real job. Worked on ERP systems for Ferrari, Boeing, Triumph. At 22, his code was processing real transactions for a car company whose cars he'd only seen in magazines. Use for: automotive, manufacturing, enterprise, Fortune 500 companies
• The Ferrari moment: first API integration he built processed a real Ferrari transaction. That's when software stopped being academic and became real. Use for: any company where "your code runs real business" resonates

Grad School & US Life (Michigan):

• Moved from Hyderabad to Southfield, MI for MS at Lawrence Tech (2024-2025)
• Culture shock of moving to the US, Michigan winters, figuring out everything from scratch
• GRA building multi-agent AI systems — the first time a CrewAI pipeline ran end-to-end autonomously felt like watching something think. Use for: AI companies, automation, agent-based systems
• Built LuffyGPT (a GPT from scratch, named after the anime character) — trained on dual RTX 3090s. Use for: LLM companies, AI research, GPU/compute companies
• Participated in Amazon Nova AI Challenge and RSNA Pneumonia Detection Challenge. Use for: Amazon, medical AI, radiology, competition-driven companies
• Currently job hunting on F-1 OPT — the urgency of needing sponsorship adds real stakes to every application

Interests & Personality:

• Anime fan (hence LuffyGPT), cricket, gaming
• Tinkerer — always building side projects, from IoT to AI agents to resume tools
• Moved countries alone at 22 for grad school
• Thinks in systems — likes understanding how things connect end-to-end

Personal Story Selection

1. Check references/personal-stories.md for an existing story for this exact company
2. If a match exists — use it
3. If no match — use the Personal Context above to generate a believable, specific connection:
   • Find the angle from Harsha's real life that most naturally connects to what this company does
   • Make it specific: a product, a moment, a year, a feeling. Not "I've always been interested in X"
   • Each cover letter must use a DIFFERENT angle — don't repeat the same story across companies
   • Ground it in a real life event from the context above, then bridge to the company
4. Save any good generated stories to references/personal-stories.md for reuse with that company

Story generation rules:

• Pick the ONE angle from the context above that fits most naturally. Don't force connections
• Add sensory details: where he was, who was around, what it felt like
• Keep it 2-3 sentences max in the hook. It's an opening, not a memoir
• If the company's domain genuinely doesn't connect to anything above, THEN ask Harsha

Cover Letter Structure

4 paragraphs. 250-400 words total. Plain text prose. No bullets, no headers, no markdown in the letter body.

Paragraph 1 — The Hook (3-4 sentences)

Open with the personal story. A real memory, a specific moment. Name the product, the year, the feeling. Then bridge naturally to the role.

Do NOT open with:

• "I am writing to apply for..."
• "I am excited to express my interest..."
• "Dear Hiring Manager, I am reaching out regarding..."
• Any variation of "I was thrilled to see this posting"

Paragraph 2 — Value Proposition (3-4 sentences)

Pick ONE achievement from GRA (LTU) or Infor that directly maps to the JD's core requirement. Expand it into a short narrative with a number. Show what you built, why it mattered, and the result.

Paragraph 3 — Technical Alignment (3-4 sentences)

Map skills to their JD requirements. Be specific — name their technologies and show you have them. Add a second proof point from a project. Don't list skills; show how you used them.

Paragraph 4 — Close (2-3 sentences)

State availability clearly. Mention willingness to relocate if the role requires it. End with a confident forward-looking sentence.

Always close with:

Sincerely,
V Harsha Vardhan Yellela
[e-mail] | [phone]
linkedin.com/in/har5ha-7663 | github.com/HAR5HA-7663

Writing Rules — Anti-AI Patterns (Non-Negotiable)

NEVER use:

Pattern	Examples to Avoid
Inflated significance	"pivotal role", "testament to", "vital", "crucial", "key role", "underscores"
AI vocabulary	"delve", "foster", "tapestry", "landscape", "pivotal", "showcase", "garner", "interplay", "intricate", "enhance", "enduring"
Promotional language	"vibrant", "groundbreaking", "renowned", "breathtaking", "cutting-edge", "innovative approach", "dynamic team"
Copula avoidance	"serves as", "stands as", "marks", "boasts", "features" — just say "is"
Negative parallelisms	"not only... but also", "it's not just X, it's Y"
Rule of three	Forcing ideas into groups of three
Synonym cycling	"role... position... opportunity... posting" — pick one word, repeat it
Em dash overuse	Use commas or periods instead
Vague flattery	"your company's innovative approach", "industry-leading solutions"
Filler phrases	"in order to", "due to the fact that", "it is important to note"
Excessive hedging	"could potentially", "I believe I might be able to"
Generic conclusions	"exciting times ahead", "I look forward to the opportunity"
Sycophantic tone	"Great company!", "I've always admired...", "Your team is amazing"
Superficial -ing	"highlighting", "showcasing", "emphasizing", "fostering", "cultivating"

ALWAYS do:

• Use contractions naturally (I'm, I've, don't, that's, it's)
• Vary sentence length — mix short punchy sentences with longer ones
• Be specific: product names, numbers, dates, team names, technologies
• Use first person naturally and unapologetically
• Have opinions — "I prefer building backend systems because..." is good
• Use straight quotes, not curly
• Write like you're talking to a smart person, not writing a formal letter
• Let some personality through — a small aside, a genuine reaction, a specific detail only you would know
• Keep it conversational but professional

Self-Check Before Output

After drafting, ask yourself: "If I read this out loud, would it sound like a person talking or a template being filled in?" If template — rewrite the stiff parts.

Output

1. Write the cover letter to /home/har5ha/Desktop/resume/cover_letters/{Company}_{Role}_Cover_Letter.txt
   • Use short role abbreviation: MLE, SDE, DevOps, AI_Engineer, etc.
   • Example: Electronic_Arts_SDE_Cover_Letter.txt
2. Display the full cover letter text to the user
3. Tell user the file path

Examples

BAD opening (AI-sounding):

"Dear Hiring Manager, I am writing to express my enthusiasm for the Software Engineer position at Electronic Arts. With my strong background in software development and passion for innovation, I believe I would be a valuable addition to your dynamic team. Your company's groundbreaking approach to gaming has always inspired me, and I am excited about the opportunity to contribute to your continued success."

GOOD opening (personal, human):

"I still remember sitting cross-legged on the floor in front of our old CRT TV, playing EA Cricket 2001 with my cousins during summer break. I couldn't have been older than seven or eight, but I was already wondering how the bowler's arm knew when to release the ball. That curiosity, how does software make something feel real, is basically the reason I became an engineer. So when I saw the Software Engineer role at EA, it felt like the kind of full-circle moment you don't ignore."

BAD skills paragraph (generic):

"I bring a diverse skill set encompassing Python, AWS, and machine learning. My experience spans multiple domains, and I have consistently demonstrated the ability to deliver high-quality solutions. I am confident that my technical expertise and collaborative mindset would make me an excellent fit for your team."

GOOD skills paragraph (specific, grounded):

"At Lawrence Tech, I built multi-agent pipelines with CrewAI and LangChain that automated our research team's literature review process. What used to take a week of manual work now runs in about 30 minutes. I deployed the whole thing on AWS EKS with OpenSearch handling semantic search across 10,000+ documents. The system is still running in production, which I'm honestly a little proud of."