Skills-md category
Productivity Skills-md for AI Agents, Claude & Codex
Productivity skills-md are SKILL.md files that teach Claude, Codex and other AI agents to do everyday work to a professional standard: finding jobs and writing CVs, planning, email, meetings, workplace writing, research, notes and decisions. Each skill is a plain Markdown file with a clear workflow, rules the agent must follow and checks it runs before answering. Copy one, download it, or take a whole set.
Jobs and Careers is the first live set. Browse its published skills below, open any skill for the full file, or copy and download it straight from this list.
Daily Job Opportunities
Run a daily search for new roles that fit the user: three qualifying routes (remote, senior home-country roles at strong employers, sponsored roles abroad), an affordability gate, ATS-first searching with verification on the employer's own page, freshness and repeat logs, scam checks, ranking, compact or detailed lists and a scheduled-task setup.
Job Fit Evaluation
Evaluate one job posting before writing a CV: deal-breaker gates for work rights, location, level, pay and hard requirements, an evidence map against the career profile, stretch, fit or overqualified level, legitimacy checks, weighted scoring only after the gates, A to F grades with clear decisions and an application plan.
CV Writer
Write the strongest truthful CV for a specific job from the user's own career profile: full JD analysis, an evidence map tying every claim to a profile line, positioning, tailored summary, skills and bullets, careful title adaptation, regional conventions, ATS-aware but human-first writing, optional cover letter, gap and risk notes and a researched 70th to 90th percentile salary range.
Cover Letter Writing
Write cover letters that earn their place: the employer's real need, an angle (problem, domain, transition, referral or product), two or three pieces of real evidence from the user's career profile, light verified company research, a 3 to 5-paragraph structure, short and email versions, variants for career changes, referrals and gaps, and regional norms.
LinkedIn Profile Optimiser
Optimise a LinkedIn profile for recruiter search and first impressions: a clear target, a keyword set from real postings, headline and About formulas, achievement-led experience, ordered skills, Featured proof, recommendations, photo, banner, URL and location basics, safe Open to Work settings and consistency with the CV.
Networking and Referral Outreach
Plan and write outreach that gets replies: choosing contacts from former colleagues to hiring managers and alumni, brief research, one clear ask per message, short LinkedIn and email templates, a forwardable referral note, one polite follow-up, thank-yous and a contact log.
Remote and International Jobs
Find and win remote and international work the user is eligible for: decoding eligibility wording, classifying roles as eligible, ask first or not eligible, employee, EOR, contractor and relocation setups, recruiter questions, like-for-like pay comparisons, tax and contract questions for professionals, visa routes via official sources and remote-job scam checks.
No skill matches that search. Clear it or choose All.
What productivity skills do for an AI agent
A general AI model can write an email or summarise a document, but left alone it works like a capable stranger who knows nothing about you. Ask it for a CV and it invents a polished summary full of phrases recruiters ignore. Ask it to triage an inbox and it sorts by guesswork. Ask it for a job list and it returns stale adverts from job boards, some of them repeats, some of them scams.
These skills change that in three ways. First, they work from files you own and keep: a career profile, a job search profile, a task list, inbox rules or a notes glossary. The agent reads them fresh each time, so it uses your facts and your vocabulary instead of guessing. Second, they set firm rules where mistakes are costly: never invent a CV claim, never repeat a job already sent, never send an email or submit an application without your say-so. Third, they end with checks, so the agent tests its own work before you see it.
The result is an assistant that behaves like an experienced colleague: specific, careful with facts, honest about gaps, and quick.
The 10 productivity skill sets
Each set will have 13 skills, installed together or one at a time. Jobs and Careers, Planning and Task Management, Email and Inbox, Meetings and Calendar, Workplace Writing and Communication and Presentations are live; the others follow in this order.
01. Jobs and Careers
Live now. Job search strategy, a master career profile, daily job opportunities, job fit evaluation, CV writing, cover letters, LinkedIn profiles, networking, application tracking, interview preparation, salary and offer negotiation, career change and remote international jobs.
02. Planning and Task Management
Live now. Daily and weekly planning, prioritisation, task capture, weekly reviews, project breakdown and personal goals, built around a task file the agent keeps up to date.
03. Email and Inbox
Live now. Inbox triage into urgent, action, read and archive, reply drafts, follow-ups, inbox rules and newsletter clean-up, with the agent drafting and you sending.
04. Meetings and Calendar
Live now. Meeting prep briefs, agendas, notes turned into actions, follow-up emails, calendar audits and scheduling across time zones.
05. Workplace Writing
Live now. Memos, proposals, reports, SOPs and decision documents, written in stages with a fresh-reader test before anyone else sees them.
06. Communication and Presentations
Live now. Status updates, stakeholder messages, feedback, difficult conversations and slides and talks that make one clear point.
07. Research and Learning
Live now. Deep research with source checks and confidence levels, recent-discussion research, summaries, reading notes and learning plans.
08. Knowledge Management
Planned. Second-brain systems, Obsidian and Notion notes, personal wikis, glossaries and decision logs that agents can read and update.
09. Thinking and Decisions
Planned. Problem framing, decision frameworks, pre-mortems, trade-off analysis and brainstorming that avoids generic answers.
10. AI Workflows and Automation
Live now. Prompting, building your own SKILL.md files, personal automations and scheduled agent tasks that run while you work.
Which productivity skills to install first
Start with the set that saves you the most time this month.
Job seekers: install Daily Job Opportunities and CV Writer together. One finds fresh, verified roles every morning; the other turns the ones you pick into tailored, truthful CVs.
Managers and busy professionals: Email and Inbox, then Meetings and Calendar, then Planning and Task Management.
Researchers, writers and consultants: Research and Learning, Workplace Writing and Knowledge Management.
Founders and freelancers: Planning and Task Management, Thinking and Decisions, and AI Workflows and Automation.
A worked example: from job list to tailored CV
A job seeker saves two files: a job search profile listing target titles, remote regions, a minimum salary and platforms to skip, and a career profile holding every role, result and tool they can prove. They set a daily task in their agent that runs the Daily Job Opportunities skill at 7:30.
Each morning the agent reads the shortlist log, searches Greenhouse, Lever, Ashby and Workable directly, opens every role on the employer's own page, drops repeats, old postings and scams, removes roles that fail the deal-breakers, ranks the rest and adds a salary estimate to each line. The user replies with two numbers. The CV Writer skill maps each job description to lines in the career profile, writes a CV and cover letter for each, and reports gaps and facts to confirm before anything is sent.
How these skills are written and checked
Each skill starts from work I do myself, then from research into the best public skills and the evidence behind them. SkillsBench, the first benchmark of agent skills, found curated skills raised task success by 16.2 percentage points on average, while skills an agent wrote for itself gave no benefit, and focused skills beat long documentation. So every skill here is human-written, focused on one job and built around a clear workflow.
The job skills also borrow proven patterns from open-source tools such as career-ops: one master career file, every claim mapped to evidence, deal-breakers checked before scoring and a human approving every application. Claims about hiring are checked against current research, such as recruiter studies showing most applicant tracking systems do not auto-reject CVs. Every file is written in British English, credits its sources and is free to use under CC BY 4.0.
Frequently asked questions
What is a productivity SKILL.md file?
A Markdown file with a name, a description and step-by-step instructions that teaches an AI agent how to do one productivity job well, such as writing a CV or finding new job openings.
Which AI agents can use these skills?
Claude Code, Claude, Codex, Cursor, Gemini CLI and other tools that support the open Agent Skills format.
Do the job skills apply for jobs automatically?
No. They find roles and write documents; you choose what to apply for, and nothing is submitted without your explicit say-so.
Will the CV Writer invent experience?
No. Every claim must map to a line in your own career profile or a fact you confirm, and gaps are reported rather than hidden.
How do I install a skill?
Download the SKILL.md file and place it in a folder with the same name in your agent's skills directory, such as ~/.claude/skills/ for Claude Code or .agents/skills/ for Codex.
Are the skills free?
Yes. You can use and adapt them under CC BY 4.0 as long as you credit Afeez Adebayo and link to the source page.
Sources
- Agent Skills specification
- SkillsBench: Benchmarking How Well Agent Skills Work Across Diverse Tasks
- career-ops on GitHub
- Enhancv ATS research
Checked on 12 October 2026. See every category on the Skills-md hub and the Marketing skills.
