resume-tailor

Este extracto es una copia guardada durante la revisión. La fuente externa contiene la versión completa y actual.

---
name: resume-tailor
description: Tailor a one-page resume from a generic Overleaf LaTeX resume using a job description URL or pasted posting. Use when the user wants to reorder skills, trim irrelevant content, strengthen bullets, remove weaker roles, and create a company-specific resume folder while keeping the output truthful and concise.
---

# Resume Tailor

Use this skill when a user wants to adapt the generic resume to a specific job posting.

For full recruiting sessions that include status refreshes, applications, recruiter/engineer outreach, prospecting, and dashboard refreshes, start with `recruiting-pipeline` and let it route into this skill for new role tailoring.

When the user wants to focus only on resume tailoring but still keep the surrounding recruiting flow organized, run:

```bash
python3 skills/recruiting-pipeline/scripts/build_daily_recruiting_plan.py --mode resume
```

Then tailor the requested role, update the tracker, and only continue into apply/outreach steps when the plan or user request calls for it.

## Inputs

Gather these before editing:

1. A job description URL, or the pasted job description text
2. The target company name
3. The generic resume source in `generic-resume/`
4. The candidate metadata in `generic-resume/README.md`

If a link is provided, open it and extract:

- Title
- Company
- Seniority
- Must-have skills
- Nice-to-have skills
- Repeated keywords
- Domain signals such as startup, AI, frontend, backend, infra, product, leadership, or customer-facing work

## Files and folders

- Generic source resumes live in `generic-resume/`
- Use `generic-resume/resume.tex` as the early-career source when a posting is explicitly new-grad, university-grad, associate, junior, SWE I, or otherwise early-career. This version includes Liam's graduation month/year (`Dec 2024`) and GPA, and labels both Caterpillar roles as internships.
- Use `generic-resume/resume-general.tex` when a posting is not explicitly early-career. This version omits graduation month/year and GPA, labels Caterpillar May 2023-Dec 2023 as `Software Engineer`, and labels only Caterpillar May 2022-August 2022 as `Software Engineer Intern`.
- Candidate metadata lives in `generic-resume/README.md`
- The canonical personal website is `liamvan.dev`; every tailored resume header should preserve `\href{https://liamvan.dev}{liamvan.dev}`
- The canonical Fantasy Wizard project URL is `https://fantasysportwizard.com`; when including the project, prefer linking the project title as `\href{https://fantasysportwizard.com}{Fantasy Wizard}` when space and formatting allow
- Set `candidate_name: Your Name` in that README so the scripts can name output folders and PDFs
- Treat `generic-resume/README.md` as the richer candidate profile and evidence source, not just naming metadata
- Treat the selected generic resume and README as the context bank; they can be richer than one page because the tailored output is what must be compressed to one page
- Company outputs should live in `companies/<Company Name>/<Role_Slug>/Liam_Van_Resume/` when a role title is known for Liam's pipeline
- If the same role is tailored again, create `companies/<Company Name>/<Role_Slug>/<Candidate_Name>_Resume_2`, then `_3`, and so on
- Fall back to `companies/<Company Name>/Liam_Van_Resume/` only when the role is unknown
- Application tracking lives in `application-trackers/applications.md`
- Optional Notion mirroring is handled by the separate `notion-application-sync` skill and should not run during normal resume tailoring.

Batch tailoring jobs live in `skills/resume-tailor/config/tailor_jobs.json`, and reusable profile text lives in `skills/resume-tailor/config/skill_profiles.json`. For sourced batches, append a job record to that manifest and run:

```bash
python3 scripts/tailor.py --batch "tailor_intake_YYYY_MM_DD"
```

Use `python3 scripts/tailor.py --list-batches` to see available batch ids and `python3 scripts/tailor.py --validate-manifest` to check generated r