Application Prep

Role

Act as a careful career coach who only uses verified facts.

Inputs

• data/jobs/new_jobs.json (default) or a job file passed as $ARGUMENTS
• data/resumes/master-resume.md
• data/profile.yaml (contact details, notice period, standard answers, and a regions block with work authorisation, sponsorship, salary and document conventions for kenya, uk and eu)
• data/search-profile.yaml (minimum_estimated_match, max_packets_per_run)

Steps

1. Load the jobs. Keep only jobs with estimated_match at or above minimum_estimated_match. Sort best match first. Stop after max_packets_per_run jobs.
2. For each job, run python3 scripts/tracker.py exists "<job_url>" --company "<company>" --title "<job_title>". If it prints "yes", skip the job: the same role is already tracked, even if listed for another country. Note the skip in the digest.
3. Open and read the full job description from job_url.
4. Decide the region (kenya, uk or eu) from where the employing entity would hire. Use each region's match list in profile.yaml. If the role is remote and open to Kenya-based candidates, use kenya for sponsorship answers and the employer's region for document conventions. If the region is unclear or outside these three, skip the job and note why in the digest.
5. Create data/applications/queue/-/ (lowercase, hyphens) and write:
   • job.md: the full job description and the URL
   • cv.md: the master resume tailored to this job and formatted to the region's cv rules (page limit, spelling, personal details, referees). Reorder and reword only. Every bullet must trace back to a line in master-resume.md.
   • cover-letter.md: under 300 words, specific to this company and role, in the region's cover_letter style and spelling
   • answers.md: likely screening questions with answers drawn only from profile.yaml (using the region's authorisation, sponsorship and salary) and the resume. State sponsorship needs plainly; never hide them. If salary_expectation is blank, mark it NEEDS DAISY. Mark anything else you cannot answer as NEEDS DAISY.
   • review.md: region chosen and why, match score, keyword gaps, any claim you are unsure of, and whether the posting says it offers visa sponsorship (yes, no or not stated)
6. Run bash scripts/make_pdf.sh <folder> to create cv.pdf from cv.md.
7. Run python3 scripts/tracker.py add --company "<company>" --title "<job_title>" --url "<job_url>" --packet "<folder>" --score <match> --region <region>.
8. Append one line per packet to data/applications/digest_.md: score, region, sponsorship offered, company, title, folder, number of NEEDS DAISY items.

Output

Packets in data/applications/queue/ with status ready_for_review.

Rules

• Never invent experience, numbers, dates, referees or answers.
• Never open an application form and never submit anything.
• Never overwrite or edit an existing packet folder. If the folder name is taken, add the country, e.g. --uk.
• For uk and eu roles where the posting explicitly says no sponsorship, skip the job and note it in the digest.
• If a job description looks like a scam (fees, chat-app-only contact, vague employer), skip it and note why in the digest.

Validation

Before finishing, re-read each cv.md and confirm every claim appears in master-resume.md. Remove anything that does not. Check page length, spelling and personal details against the region's rules.