Productivity
Productivity agent skills automate everyday work: task management, email drafting, calendars, note-taking systems, and personal workflows. Each is a small SKILL.md file your AI agent applies automatically when your request matches what the skill does.
-
wolf123450 Skill Claude Code Token OptimizationUse when the user wants to cut Claude Code token/context spend, complains sessions are getting expensive, or asks to set up session-hygiene and subagent-delegation discipline — including replicating this setup on a new machine or in a fresh environment.
-
yezhwi Bundle Engineering HarnessUse when starting or resuming a task in a repo with .harness/ state, when unsure which harness phase comes next, or when asked to run the Engineering Harness loop end-to-end. Orchestrates the fixed state machine; MUST NOT implement business code itself.
-
yezhwi Skill Quality GateUse when task state is GATING and a go/no-go decision is needed before CONVERGED. Runs `harness gate` and follows `DECISION:`; MUST NOT assess quality by judgement.
-
yezhwi Skill Task ContractTurn a User Requirement into a Task Contract: Acceptance Criteria, Requirements, Invariants, Risks, Verification Plan. Produces .harness/requirements.yaml and .harness/invariants.yaml and advances task state CLASSIFIED -> SPECIFYING -> PLANNED. This skill MUST NOT implement business code.
-
yezhwi Skill Collect EvidenceUse when entering VERIFYING or before GATING and fresh evidence is required. Decides which evidence types the task needs, then runs collect_evidence.py; MUST NOT judge pass/fail itself.
-
yezhwi Skill Adversarial ReviewAdversarial review of a Task Contract: construct failure scenarios, attempt to violate Requirements and Invariants. Produces .harness/findings/FND-nnn.yaml files with status=PROPOSED only. This skill MUST NOT confirm bugs, review code style, or suggest refactors.
-
misssophiesterling Skill PinpointTurn a vague task into a pinpointed brief before searching, and map a codebase once so later tasks start in the right file. Use when a request has no location or expected result ("fix my site", "debug this", "the button doesn't work"), or when asked to map where things live in a repo.
-
misssophiesterling Skill Workflow OptimizerMake a recurring job cheaper. Measure a baseline, test cheaper variants against the same quality check, keep the winner as a skill. Use when the user wants a repeat workflow (reports, content batches, triage, builds) to cost less ("optimize this workflow", "get this under 1% of my usage").
-
ehteshambuildagents Skill Bl WeekPlan the week on LinkedIn - what to post, when to post it, and who to engage with. Use when the user says "plan my week", "what should I post", "content calendar", "I have nothing to post about", or wants a posting schedule and an engagement list.
-
mypropper Skill Extract Contract DatesPull every date, deadline and notice window out of an agreement and turn them into a dated list — effective date, term and expiry, auto-renewal with the notice deadline that precedes it, termination rights, cure periods, and payment or delivery milestones. Use when someone asks when a contract renews or expires, what the notice period is, what they need to put in a calendar, how long they have to get out of something, or wants the deadlines pulled out of a stack of contracts.
-
vsevolodbazhan Skill Jira Create TicketWhen user asks to create a task or a ticket in Jira.
-
zime-skills Skill Follow UpDrafts the follow-up email after one call. Recaps what was agreed and states the next step, using only what was actually said. Always a draft for you to review before sending.
-
zime-skills Skill Review CallEverything about one call or meeting, at whatever depth you need: the record, a structured recap, or the exact words. Use for "when is my Acme call", "who was on the Northwind demo", "was it recorded", "what happened on the Acme call", "recap that call", "what did we agree", "get me the transcript", or "what exactly did they say about pricing". Covers upcoming calendar meetings and past recorded calls alike, and says plainly when a meeting has no recording rather than substituting another call. Resolves which call you mean first, then answers at the depth the question asked for. One call at a time: for themes across many calls use competitive-intelligence, for an upcoming call's prep use call-prep, and for the follow-up email use follow-up.
-
ivelten Skill Haskell Quality GatesConfigure Haskell formatting, linting, performance defaults, pre-commit hooks, and CI checks. Covers ormolu, hlint, cabal-gild, the .pre-commit-config.yaml template, CI workflow steps, strict-by-default field policy, Text/ByteString/Vector/Map preferences, profiling flags, and the "before declaring work done" checklist. Use when wiring up linting/formatting automation, setting up CI for a Haskell project, debugging space leaks, or running the final pre-commit checks.
-
antsteyer Skill PrFull PR creation workflow: generate description, create GitHub PR with Jira link when branch starts with PROD-, assign it to the current GitHub user, and transition the Jira ticket to "Ready for review". Use when the user says "create a PR", "open a PR", "push and open a PR", or "/pr".
-
guysilvere Bundle Pwa CadrageCadrage complet d'un projet PWA Agence Bulles — définition du problème, stress-test de l'idée, naming, blueprint, stack, monétisation, copywriting landing et feuille de route. Charger au démarrage de tout nouveau projet (phase 0–1 du workflow) ou avant une refonte majeure.
-
pixee Skill Pixee WorkflowList, create, update, run, and delete Pixee workflows on a repository with partial-update semantics across event kinds.
-
roboco-io Bundle IntentIntent Document(INTENT.md) 생성 및 관리 스킬. 프로젝트의 의도를 Why/What/Not/Learnings로 문서화하고, 탐구와 학습을 통해 의도를 진화시키는 생명주기를 관리한다. 반드시 사용해야 하는 경우: "intent document", "INTENT.md", "intent engineering", "프로젝트 의도 정리", "왜 만드는지 정리", "프로젝트 목적 문서화", "What/Why/Not 정리", "프로젝트 방향 잡기", "피벗", "프로젝트 종료 판단" 등을 언급하거나, 새 프로젝트를 시작하면서 방향을 잡고 싶다고 할 때. 기존 INTENT.md를 수정하거나 학습 결과를 기록하고 싶을 때도 이 스킬을 사용한다.
-
ftrout Bundle Skill TemplateReplace this with what the skill does AND when to use it. List the phrases, artifacts, and situations that should trigger it, including cases where the user does not name the task explicitly. Keep under 1024 characters. Be a little pushy; Claude under-triggers skills.
-
candy14432 Bundle Usability Testing規劃、執行與整理可用性測試,用於發現操作障礙、驗證流程與優先排序改善項目。當任務涉及使用者測試、task flow 驗證、原型評估、可用性問題盤點或設計方案比較時使用。
-
jtydhr88 Bundle Lw WorkflowEntry point and router for the lyric writing library (作词工作流与路由). The two workflows that fork at the very first question - music-first versus lyrics-first - and their different stage tables, which skill to load for which task, the full-library boundary table, the LYR-SPEC gate rules, the bidirectional interface with the arrangement library, and the three testing devices. Load this FIRST for any lyric writing request, before deciding which other skill to read. Use when starting a lyric, when fitting words to an existing melody or a melody to existing words, when unsure which skill owns a question, when a draft fails lint, or when handing work to another library. 作词、歌词、填词、工作流、路由、该读哪个 skill、LYR-SPEC、先曲后词、先词后曲。
-
shreyp087 Bundle AI Eval LabEvaluate an AI hackathon feature against a simple baseline using task cases, failure labels, latency, and cost before making reliability claims.
-
kroshank Skill Meetings Minutes OrganizerTurn messy meeting notes into clean minutes with decisions, tasks, owners, and deadlines. Use when the user pastes raw notes, a transcript, or a call dump and wants an acta or action list.
-
mzvonar Bundle Lessonslessons — capture durable insights during ad-hoc work, promote them later
-
mzvonar Skill Update SkillChange a skill that was installed from the claude-skills-public marketplace (any `/dev-tools:*`, `/workflow:*`, `/next-js:*`, `/prisma:*`, `/describe-changes:*`, `/refdiff:*`, `/svc:*` skill) the right way — upstream in a sibling checkout, validated, version-bumped, then rolled out with a plugin update. Use whenever the user asks to update, fix, improve, extend, or reword one of these skills, says a skill "is wrong" or "should also…", or you are about to edit a file under ~/.claude/plugins/cache/ or copy a plugin skill into .claude/skills/. Never edit the cached copy and never vendor it into the repo.
-
kroshank Skill Study Session PlannerBuild a realistic study calendar from available days, subjects, and difficulty, including breaks. Use when the user wants a study timetable, revision calendar, or session plan before exams.
-
kroshank Skill Professional Email WriterProfessional Email Writer
-
drdustinedwards Bundle Cloudflare Email ServiceImplement or troubleshoot Cloudflare Email Sending and Email Routing integrations and their delivery configuration.
-
drdustinedwards Skill Reproduce Before FixingReproduce a reported defect before changing any code to fix it. Load when a task says a guard, test, check or gate is broken, misses something, or passes when it should fail, and before writing the fix.
-
g-tavares14 Skill Git Workflow And VersioningStructures git workflow practices. Use when making any code change. Use when committing, branching, resolving conflicts, splitting uncommitted work in a messy working tree into clean atomic commits, opening or reviewing a pull request (PR), pushing to a remote, or when you need to organize work across multiple parallel streams. Use when cutting a release, choosing a semantic version bump, tagging, or writing a changelog.
-
808enzo Bundle Email CopyDecide what one email says, what it promises, what it asks for, and how you know the words were the reason it worked. Use when a message has to be written or rewritten, when a subject line has to be chosen, when nobody can say in one sentence what a send is for, when a message carries two jobs at once, when clicks land on navigation instead of the offer, when a promise made at signup has to be kept in words, when a substitution can come out empty, or when a draft is about to go out and somebody has to decide whether it can. Covers the message brief, inherited promises, claim discipline, the sender name, subject and preheader, body order, the single call to action, fallback wording, link roles, and the review that blocks a send. Not who receives the message, not when it fires, not the economics of the offer, not the template, and not a list of spam words.
-
808enzo Bundle Offer DesignDecide what you promise a customer in exchange for a purchase, down to the form of the benefit, how deep it goes, what it is conditional on, how long it lives, what carries it, and what happens when several promises land on one order. Use when somebody has decided to "send a discount" and nobody has said how much, when a promo code leaks, when two promotions stack into a loss, when a code stops working while the message promising it is still in inboxes, when you are pricing a game or prize mechanic, or when you want to cut discounting without losing the base. Covers benefit form, depth against margin, entry thresholds, redemption windows, exclusions, unique against shared carriers, issuance limits, prize funds, arbitration between offers, and the fate of a promise you can no longer honor. Not loyalty program construction, not the promotional calendar, not which mechanic sends the message.
-
808enzo Bundle Email ProgramRun email as a standing program rather than as one-off sends. Use when planning what goes out and to whom, setting a sending rhythm per engagement tier, building a slot calendar, deciding who is excluded from a given send, or diagnosing a channel whose returns are falling. Covers program design, recipient selection, and recovery. Not for writing or designing a single email, not for authentication and inbox placement, not for event-triggered flows, and not for the cross-channel contact cap.
-
808enzo Bundle Promo CalendarDecide what happens when in the year, and what a peak does to the weeks around it. Use when building or rebuilding the promotional calendar for a cycle, working out whether a season is real or self made, deciding which occasions earn a place and which do not, preparing and running a peak window, placing windows so they do not collide with each other or with promises already issued, filling a trough, or explaining why a sale that looked good left the year smaller. Covers the index of a period and the separation of season from trend, the admission test for an occasion, the window register, the issuance cutoff, lead time and the message and channel ladders, the freeze window, the extended window a verdict is read on, and the three directions demand can be shifted. Not discount depth, not the channel's own calendar, not the frequency cap, not copy or design.
-
jtydhr88 Bundle Mc WorkflowEntry point and router for the music composition library (作曲编曲工作流与路由). The S0-S7 stage table from brief to delivered audio, which skill to load for which task, the full-library boundary table that prevents two skills claiming the same decision, the ARR-SPEC gate rules, and the three testing devices (routing, text-layer A/B, compliance rate). Load this FIRST for any request that will produce or change an ARR-SPEC (a whole track, a section rewrite, a generation run). A single knowledge question or a single-part task goes straight to the owning skill without this one. Use when starting a new track, when unsure which skill owns a question, when a spec fails lint, when a generated result sounds wrong and you need the diagnostic path, or when handing work to another library. 作曲、编曲、工作流、路由、该读哪个 skill、ARR-SPEC、出歌流程。
-
808enzo Bundle Subscription RetentionKeep a paid subscription from ending without the person's decision, and end it cleanly when the person decides. Use when a renewal charge fails and nobody knows how long the person keeps access or when they count as gone, when the cancel screen holds a discount and a legal team asks where the cancel button is, when annual renewals produce refund requests and "cancelled recurring" disputes, when subscriptions billed through an app store get your dunning emails, or when a paused subscription never resumes. Covers the notice calendar, the recovery case after a declined charge (decline classes, retry window, access, deadline), and the exit request with one alternative, one offer, the effective date and the pause. Not the general form of churn, not the signals before a lapse or the win-back after it, not the contract renewal handled by an account manager, not the trial's activation, not the depth of a discount, not the class of a service message.
Frequently asked questions
What are Productivity agent skills?
Productivity agent skills automate everyday work: task management, email drafting, calendars, note-taking systems, and personal workflows. Each is a small SKILL.md file your AI agent applies automatically when your request matches what the skill does.
Which Productivity skills are most installed?
Popular Productivity skills on SkillMD right now include claude-code-token-optimization, engineering-harness, quality-gate. Rankings shift as installs change; sort this page by "Most installs" for the live list.
Do Productivity skills work with Claude Code and Cursor?
Yes. Every skill here ships as a SKILL.md file, an open format that works in Claude Code, Claude.ai, Cursor, Codex, Windsurf, and 60+ other agents. Install one with npx skillmds@latest add <owner>/<name>, or copy the file into your agent's skills directory.