splash0047/resume-generator / .agents/skills/candidate-positioning
Candidate Positioning
VerifiedDetermines the candidate's primary role identity before any JD analysis. Classifies into role families, selects core tech identity, ranks projects by role relevance, and outputs a Candidate Position Brief consumed by all downstream stages. This is Phase -1 of the pipeline.
Candidate Positioning
When to Use This Skill
Use this skill as Phase -1 — the very first step — before JD analysis, before bullet generation, before anything. This skill answers the most important question:
"What kind of candidate are you?"
Trigger phrases: "position me for", "what role should I target", "generate resume", "create resume variant", any resume generation request
This skill runs once per candidate profile, then the output is reused across all JD-specific tailoring.
Core Principle
A recruiter must identify your role identity within 6 seconds. A resume that says "I know 30 technologies" communicates nothing. A resume that says "I'm an AI/ML engineer who builds production ML systems with Python and FastAPI" communicates everything.
Positioning happens BEFORE keyword optimization. You cannot optimize what you haven't positioned.
Step 1: Profile Analysis
Analyze the candidate's full profile (INPUT 2) across these dimensions:
Experience Signals
| Signal | Weight | What to Extract |
|---|---|---|
| Job titles | High | Direct role identity (e.g., "AI Intern" → AI/ML signal) |
| Technologies used at work | High | Production-level tech stack |
| Project domains | Medium | What problems they solve (ML, web apps, CLI tools, infra) |
| Certifications | Medium | Industry validation direction (Azure AI, AWS, etc.) |
| Education focus | Low | Coursework alignment |
Project Technology Distribution
Count how many projects use each technology category:
- AI/ML: ML models, LLMs, RAG, agents, vector databases, anomaly detection
- Backend: REST APIs, FastAPI, Express, databases, authentication
- Frontend: React, Next.js, UI components, responsive design
- DevOps/Cloud: Docker, CI/CD, cloud deployment, infrastructure
- Data Engineering: Pipelines, ETL, data processing, streaming
- CLI/Tooling: Command-line tools, developer experience, automation
Step 2: Role Family Classification
Based on the profile analysis, classify into ONE primary and ONE secondary role family.
Available Role Families
| Role Family | Key Signals |
|---|---|
| SDE / Software Engineer | Strong DSA, multiple languages, backend/API work, system design, production software |
| AI/ML Engineer | ML models, LLMs, RAG, AI agents, data pipelines, Python-heavy, research papers |
| Full Stack Developer | React/frontend + Node/backend, MERN/MEAN, responsive UI, authentication, RBAC |
| DevOps / Cloud Engineer | Docker, Kubernetes, CI/CD, cloud certifications, infrastructure, monitoring |
| Data Engineer | ETL pipelines, Spark, Kafka, data warehouses, SQL-heavy, streaming |
Classification Rules
- Primary role = the role family with the most evidence (projects + experience + certifications combined)
- Secondary role = the next strongest, used for variant generation
- If tied: Prefer the role family that matches the candidate's most recent experience
- Never classify as 3+ role families — that's the "I know everything" anti-pattern
Classification Output
## Role Family Classification
**Primary:** AI/ML Engineer
**Evidence:** 3/6 projects are ML/AI, AI internship, Azure AI + Oracle AI certifications, LangGraph/CrewAI/RAG experience
**Secondary:** Software Engineer (Backend)
**Evidence:** FastAPI, REST APIs, Docker, Node.js/Express, database design
Step 3: Core Technical Identity
Select 8-12 core technologies that define this candidate for the primary role family. These are the technologies that MUST appear prominently in the resume.
Selection Rules
- Include: Technologies the candidate has used in ≥2 projects or ≥1 work experience
- Include: Technologies that appear in >50% of JDs for the primary role family
- Exclude: Technologies used in only 1 tutorial-level project
- Exclude: Technologies that dilute the role identity (e.g., CrewAI for SDE roles)
- Maximum 12 — fewer is stronger
Core Identity by Role Family (Examples)
For SDE:
Core: Python, C++, JavaScript, TypeScript, SQL, FastAPI, Node.js, Express, React, Docker, Git
Remove/Minimize: CrewAI, LangGraph, RAG, Neo4j, Azure OpenAI, FAISS, LangChain
For AI/ML Engineer:
Core: Python, ML, Scikit-learn, LLMs, RAG, LangGraph, FastAPI, FAISS, Elasticsearch, Docker, Git
Remove/Minimize: Next.js, Tailwind CSS, Bootstrap, Prisma ORM, GraphQL
For Full Stack:
Core: React, Next.js, Node.js, Express, TypeScript, MongoDB, MySQL, REST APIs, Docker, Git, Tailwind CSS
Remove/Minimize: CrewAI, LangGraph, FAISS, Elasticsearch, Neo4j, Isolation Forest
Step 4: Project Priority Ranking
Rank ALL candidate projects by relevance to the primary role family. Output a priority list with selection recommendations.
Ranking Criteria
| Factor | Weight |
|---|---|
| Technology overlap with core identity | 40% |
| Problem domain relevance | 25% |
| Architecture complexity | 20% |
| Deployment/demo evidence | 15% |
Output Format
## Project Priority for [Role Family]
### Include on Resume (Top 2-3)
1. **[Project Name]** — [Why it fits: covers X core technologies, demonstrates Y]
2. **[Project Name]** — [Why it fits]
3. **[Project Name]** — [Why it fits, if space permits]
### Keep in Portfolio Only
4. **[Project Name]** — [Good project but doesn't fit this role variant]
5. **[Project Name]** — [Domain mismatch for this role]
### Recommendation
Lead with [Project 1] — it covers [N] critical technologies and demonstrates [problem-solving type].
Step 5: Skill Section Strategy
Determine the Technical Skills section layout for the resume.
Rules
- Maximum 5 categories in Technical Skills
- Maximum 15 total items across all categories for fresher resumes
- Order categories by JD relevance (most relevant first)
- Order items within categories by proficiency and relevance
- Remove technologies that:
- Are not in the core identity AND not in the JD
- Were used in only one tutorial-level project
- Dilute the primary role signal
Category Templates by Role
SDE:
Languages: Python, C++, JavaScript, TypeScript, SQL
Backend: FastAPI, Node.js, Express.js, REST APIs
Frontend: React.js, HTML5, CSS3
Tools: Docker, Git, GitHub, CI/CD
Databases: MongoDB, MySQL, PostgreSQL
AI/ML:
Languages: Python, SQL, JavaScript
AI/ML: Scikit-learn, LLMs, RAG, LangGraph, CrewAI, FAISS
Backend: FastAPI, REST APIs, Elasticsearch
Cloud: Azure AI, AWS, Docker, CI/CD
Databases: MongoDB, PostgreSQL, Neo4j
Step 6: Certification Selection
Select 2-3 certifications most relevant to the primary role family.
Rules
- Maximum 3 certifications on a fresher resume
- Relevance > quantity — 2 relevant certs beat 5 irrelevant ones
- Order by: Role relevance → Issuer prestige → Recency
- Certifications not selected still exist on LinkedIn — they're not deleted, just not on THIS resume
Full Output: Candidate Position Brief
This brief is consumed by ALL downstream stages (JD Analyzer, Bullet Generator, Humanizer, Critic).
## Candidate Position Brief
**Candidate:** [Name]
**Primary Role:** [Role Family]
**Secondary Role:** [Role Family]
**Generated:** [Date]
### Core Technical Identity (8-12 technologies)
[Ordered list]
### Project Priority
1. [Project] — Include
2. [Project] — Include
3. [Project] — Include if space
4. [Project] — Portfolio only
...
### Skill Section Layout
[Category: items] format
### Certifications to Include
1. [Cert]
2. [Cert]
### Technologies to Remove/Minimize
[List with reasons]
### Candidate Positioning Statement
(1-2 sentence identity used internally, NOT placed on resume)
"[Name] is an AI/ML-focused backend engineer with internship experience building
ML systems and multiple deployed AI projects. Strongest in Python, FastAPI,
and ML pipeline development."
Rules
- This skill runs BEFORE the JD Intelligence Analyzer. The JD Analyzer then uses this brief to refine (not override) the positioning.
- One primary role per resume. If the JD clearly fits the secondary role, switch — but never try to be both simultaneously.
- The brief is a filter, not a fabrication. It selects and orders true experience. It never adds skills or projects the candidate doesn't have.
- Update the brief when the candidate profile changes (new project, new certification, new experience). Don't regenerate for every JD.
- For experienced candidates (3+ years): This skill is less critical because work experience naturally positions. For freshers, it's THE most important step.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deployment | Stars: 319k | Updated | Claude CodeCodex | Verified | Install |
| xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |