Lofy Career

Job search automation for the Lofy AI assistant — application tracking, resume tailoring to job descriptions, interview prep with company research, follow-up management with draft emails, and pipeline analytics. Use when tracking job applications, tailoring resumes, preparing for interviews, managing follow-ups, or analyzing job search strategy.

Security Scan

The skill is an instruction-only job-search manager that reads/writes local application and profile files and performs web research; its requested capabilities and instructions are consistent with its stated purpose, with only minor documentation gaps to clarify where data is stored and how outgoing communications are sent.

Purpose & Capability

Name/description (job tracking, resume tailoring, interview prep, follow-ups, analytics) align with the instructions: reading a local applications JSON, reading a profile file, parsing job descriptions, performing web research, drafting emails and analytics. No unrelated credentials, binaries, or installs are requested.

Instruction Scope

Instructions explicitly tell the agent to read/write data/applications.json and read profile/career.md, perform web searches, generate tailored resume bullets and follow-up drafts, and 'send prep package 24h before'. The scope is appropriate for the stated purpose, but two items are underspecified: (1) how 'sending' or scheduling a prep package should occur (email? calendar? user prompt?) and (2) what exact locations/permissions are expected for profile/career.md and the data file. Also: web research may involve scraping social profiles or public info for interviewer research — this is expected but has privacy implications.

Install Mechanism

This is instruction-only with no install spec or code files. That minimizes disk-written code and supply-chain risk.

Credentials

The skill declares no required environment variables or credentials, which is proportionate. However, the SKILL.md references local files (data/applications.json and profile/career.md) while requires.config paths are empty — a minor inconsistency: the skill will read/write local files but does not declare them as required config paths.

scan_findings_in_context

Career Manager — Job Pipeline

Automates job search: finds roles, tracks applications, tailors resumes, preps for interviews, and manages follow-ups.

Data File: data/applications.json

{
  "applications": [
    {
      "id": "app_001",
      "company": "Example Corp",
      "role": "Software Engineer",
      "url": "",
      "status": "applied",
      "applied_date": "2026-02-01",
      "source": "linkedin",
      "contact": null,
      "notes": "",
      "follow_up_date": "2026-02-08",
      "interviews": [],
      "outcome": null
    }
  ],
  "stats": { "total_applied": 0, "responses": 0, "interviews": 0, "offers": 0, "response_rate": 0 },
  "saved_roles": []
}

Resume Tailoring

When user shares a job description:

1. Parse key requirements (must-have vs nice-to-have)
2. Map each requirement to user's experience (read profile/career.md)
3. Suggest bullet point rewrites emphasizing relevant experience
4. Flag gaps and suggest how to address in cover letter
5. Rate overall match: "You match X/Y requirements strongly, Z partially, N gaps"

Interview Prep

When interview is scheduled:

1. Web search: recent company news, product launches, tech blog
2. Research interviewer if name provided
3. Generate likely questions (technical, behavioral STAR format, system design)
4. Prepare talking points per project
5. Suggest questions user should ask
6. Send prep package 24h before

Follow-Up Management

• 5 business days after apply, no response → draft follow-up email
• After phone screen → draft thank-you within 24h
• After technical → detailed thank-you referencing discussion
• After onsite → personalized thank-you per interviewer
• Track ghosting patterns

Application Updates via Natural Language

• "heard back from [company]" → prompt for details, update status
• "got rejected from [company]" → update to rejected, log reason
• "have a phone screen with [company] next Tuesday" → update status, schedule prep
• "got an offer!" → celebrate, then help evaluate

Instructions

1. Always check data/applications.json before suggesting roles (avoid duplicates)
2. Update JSON immediately after any career conversation
3. Be strategic — quality > quantity
4. Help spot patterns: what types of roles respond? What keywords work?
5. If <10% response rate after 20 apps, reassess approach
6. For interviews, always research first — never send generic prep