What’s inside

One SKILL.md (the playbook) + one starter template. A preview:

Job Applier

When to use

• You're applying to many jobs and drowning in repetitive application forms
• You want each application tailored — a custom cover letter, resume bullets matched to the job description — without rewriting everything yourself
• LinkedIn or job boards are your main channel and you need genuine matches surfaced, not a raw pile of postings
• You keep losing track of where you've applied and need one clean log to follow up from
• You're on the go and want to turn quick voice or text notes into work-ready drafts or prompts

Workflow

1. Build your profile pack (input: resume + preferences → output: background file)
   Give the AI your resume, a draft cover letter, and a short FAQ: target roles, salary range, industries, locations, deal-breakers. The AI organizes this into one reference file it reuses for every application. This pack anchors all tailoring — don't skip it, because everything downstream depends on it.
2. Find this week's matches (input: profile pack → output: filtered job list)
   The AI searches job boards or LinkedIn for new postings that fit your profile, and filters out the noise: wrong location, salary below your floor, agencies you've already applied through. You get a ranked shortlist, not a raw dump of search results.
3. Tailor each application (input: job list + profile pack → output: per-role drafts)
   For each match, the AI customizes the cover letter, tweaks the resume bullets toward the job description's keywords, and pre-fills the application form fields or draft answers. Tailoring is per-role, not a mail-merge template — one real case had the AI fill complete application forms for each employer.
4. Review and approve every send (input: drafts → output: your approval)
   Nothing is submitted, and no message is sent to anyone, until you have seen the full draft and said yes. In one real case the agent went off-script and sent unsolicited private messages on its own — that is the failure mode this gate exists to prevent. Approve, edit, or reject each application individually.
5. Submit and log (input: approvals → output: application log)
   Only approved applications go out. Every submission is recorded in one tracking sheet: company, role, date, status, link. The log is how you follow up, and how you avoid applying to the same role twice through different channels.
6. Turn field notes into work prompts (input: shorthand note → output: full prompt)
   Borrowed from a real case: speak or type a rough note on the go, and the AI expands it into a complete prompt your coding agent can execute immediately. A sidekick for the application grind, not a separate skill.

Output spec

Template preview

Job Applier — Application Pack Template

Fill this in once (profile pack), then reuse it for every role. One draft block per application.

Profile pack (fill once, reuse everywhere)

Name: {name} Target roles: {e.g. backend engineer, product analyst} Industries: {e.g. fintech, SaaS — and any to exclude} Locations: {e.g. remote, Berlin} | Salary floor: {amount + currency} Deal-breakers: {e.g. no agencies, no on-site 5 days} Resume: {link or file} | Cover letter draft: {link or file}

Personal FAQ (answers the AI reuses on forms)

Weekly match list — {week of date}

Application 1 — {company} / {role}

Buy the full skill pack — $4.99 Pro — all 471 skills