| name | resume-analyst |
| description | Analyzes a person's resume (PDF) and produces a structured assessment covering sentiment, skills and experience inventory, strengths, weaknesses, notable achievements, recruiter questions, and recruiter concerns. Use when the user wants to evaluate, review, score, or get feedback on a resume or CV, or when the user asks what a recruiter might think about someone's profile. |
| compatibility | Requires Python and pdfplumber (via the read-pdf-document sub-skill) |
| metadata | {"author":"franck-boullier","version":"1.1"} |
Resume Analyst
Produces a structured, multi-section assessment of a resume from a PDF file.
Prerequisite: check for the read-pdf-document skill
Before doing anything else, verify the read-pdf-document skill is available by checking whether the file ../read-pdf-document/SKILL.md exists relative to this skill's directory.
-
If the file does not exist, stop immediately and return this error to the user:
Error: The resume-analyst skill requires the read-pdf-document skill, which was not found.
Please install the read-pdf-document skill in the same skills directory before running this skill again.
-
If the file exists, read it and follow its instructions to extract the resume text.
Workflow
Task Progress:
- [ ] Step 0: Collect PDF input
- [ ] Step 1: Verify read-pdf-document skill is installed
- [ ] Step 2: Extract full text from the resume PDF
- [ ] Step 3: Analyse the extracted text
- [ ] Step 4: Produce the structured report
Step 0 — Collect PDF input
If the user has not already provided a PDF resume file, stop and ask:
"Please provide the resume PDF file you would like me to analyse."
Wait for the user to supply the file before proceeding to Step 1.
Step 1 — Verify dependency
Check for ../read-pdf-document/SKILL.md relative to this skill's location. Error out if missing (see above).
Step 2 — Extract resume text
Read the read-pdf-document skill and follow its instructions to extract all text from the supplied PDF file. That skill handles layout-preserving extraction and OCR fallback for scanned documents automatically.
If the extracted text is empty, stop and apply the edge-case guidance in the Edge Cases section below.
Step 3 — Analyse the resume
With the extracted text, perform the following analysis:
| Section | What to assess |
|---|
| Sentiment | Overall tone (positive/neutral/negative), confidence level, assertiveness, professional register, use of action verbs vs passive language |
| Skills & Experience | Hard skills, soft skills, tools/technologies, years of experience per domain, industries/roles covered |
| Strengths | Areas where the candidate clearly excels or has strong evidence to back claims |
| Weaknesses | Gaps, vague claims, inconsistencies, missing information, potential red flags |
| Notable Achievements | Quantified results, awards, promotions, unique contributions that differentiate the candidate |
| Recruiter Questions | Questions a recruiter would likely probe during a screening call, based on gaps or interesting points |
| Recruiter Concerns | Concerns a recruiter might flag: gaps in employment, short tenures, missing skills for typical roles, etc. |
Step 4 — Produce the report
Format the output using the template in references/OUTPUT-TEMPLATE.md.
Edge Cases
| Scenario | How to handle |
|---|
| Empty or unreadable PDF (no text extracted after OCR fallback) | Stop immediately. Return: "No text could be extracted from the provided PDF. The file may be corrupted, password-protected, or contain only images without OCR support. Please provide a different file." |
| Very short resume (fewer than ~200 words of extracted text) | Proceed with analysis but add a note in the report header: "The extracted resume content is unusually sparse. Some sections may have limited findings." |
| Non-English resume | Add a note at the top of the report that the resume appears to be in a language other than English. If the content is intelligible, proceed. If not, ask the user whether to continue. |
Guidelines
- Be specific: quote or paraphrase evidence from the resume to support every point.
- Be balanced: list both genuine strengths and real weaknesses — avoid being sycophantic.
- Keep each section focused: 3–7 bullet points per section is usually right; use more only when clearly warranted.
- For the Recruiter Questions section, phrase each item as a direct question a recruiter would say out loud.
- For the Recruiter Concerns section, phrase each item as a concern statement, not a question.
- If a section has nothing meaningful to report (e.g. no notable achievements found), say so explicitly — do not fabricate.
