Skills
99,951 packaged capabilities and playbooks for agents. Install any of them with one command.
All
Analyze industry sectors (industry indices, constituent stocks, fund flow rankings, historical performance) to judge sector strength and rotation. Use when the user asks about sector performance, sector rotation, sector fund flows, sector constituents, or needs sector comparisons.
Review influencer/creator agreements for content rights, exclusivity, FTC compliance (16 CFR 255), AI/Synthetic Media consent, whitelisting/paid usage, and platform-specific requirements. Covers: Instagram, TikTok, YouTube. Includes AI Influencer/Virtual Influencer guidance. Triggers: (1) 'check contract' → checklist review, (2) 'advise' / 'review' → full consultation with playbook, (3) 'generate' / 'template' → Skala template URL. Jurisdiction: New York, USA.
Set up PostHog in a user's GitHub repository from a chat, without a terminal. Wizard, PostHog's setup agent, runs in the cloud, adds the SDK and first events, and opens a pull request. Use only when you can't read or edit the user's code, which is common in web chat apps such as ChatGPT on the web or Claude on the web. Typical requests: install, add, or set up PostHog in their app or repository, or open a pull request that adds PostHog. When you're working in the user's repository, use installing-posthog-in-a-codebase instead. Covers connecting GitHub, picking the repository, starting the run, polling it, and sharing the pull request.
Use when the user types `/integrate`, `/integrate --product`, or asks to "integrate a Juspay product", "set up payments", "add payment SDK", or any variation of adding a Juspay payment product to their app or codebase. Activate for any Juspay payment integration request.
Transform vague ideas into implementation-ready specifications through structured interviewing. Use when user describes a new feature/product idea, has a problem to solve, or needs to document requirements. Produces intent.md (technical spec for code agents) and overview.md (human-friendly summary).
This skill provides comprehensive support for the writing process from ideation through revision. Use this skill when helping users write essays, articles, or creative pieces through interactive collaboration. The skill supports co-evolving outline and prose, voice-based input processing, multiple writing styles, and connection to the user's PKM system for enriched content.
Activates when the user asks to write, draft, translate, or revise Italian texts for public, editorial, blog, social media, newsletter, marketing, fiction, nonfiction, or informational use. It also activates when the user asks to humanize Italian texts, remove AI patterns, correct chatbot-like wording, capture or calibrate an author's voice based on a corpus of their texts, respond to an automatic detector that flagged human-written text as generated, or revise Italian text generated by language models.
A skill that applies writing conventions when having generative AI create or polish Japanese technical documents (book chapters, articles, explanations). It produces logical, honest Japanese technical writing and suppresses LLM-specific empty phrasing, verbosity, over-dramatization, and AI-sounding vocabulary. It can also score texts before publication. Triggers: 日本語技術文書, テクニカルライティング, 技術記事を書いて, 解説文を書いて, 技術文書を推敲, 文章を校正, AIっぽい文章を直す, AI臭を消す, 公開前チェック, write Japanese technical doc, japanese tech writing.
Produce an implementable JAX, Flax, and Optax training recipe with explicit PRNG management, pure functions, jit/vmap/pmap usage, gradient computation, Flax state handling, checkpointing, and debugging checks. Use when implementing a model in JAX/Flax, porting from PyTorch, debugging tracing/shape/NaN issues, or designing research code that uses JAX transformations.
Reference for the `just` command runner — justfile syntax and GitHub Actions integration. Use when an existing repo already has a `justfile` and tasks must be added, modified, or audited. For new mizchi-owned repos, prefer the `pkfire` skill (`Taskfile.pkl` + `pkf`) instead; justfile is kept only for compatibility with existing repos.
A Skill dedicated to the living room renovation pitfall-avoidance guide. It covers the full process of living room layout, sofas and coffee tables, TV walls, ceilings, floor tiles, feature walls, no-main-light designs, lighting and outlets, curtains, balconies, style matching, storage, projectors, and daylighting and circulation, integrating the essence of the 西安装修课堂 knowledge base in reverse order to help users create a practical and high-quality living room.
Create, read and manage KubeSense service level objectives (SLOs) over the REST API — by_count, time_slice and alert-uptime SLIs, error budgets, burn-rate alerting and grouping — and translate Datadog SLOs into equivalent KubeSense ones. Covers the exact POST /api/slo payload, the good/total query config shape it shares with alert rules, and the traps that make an SLO store cleanly and then evaluate nothing.
Smart Label Capture (Scandit `LabelCapture`) in native Android projects — extracting multiple fields (price, expiry date, serial or lot number, weight) from a label in one scan, using barcode fields plus text fields with regex patterns and pre-built definitions like price capture. Use for integration, label-definition configuration, captured-session handling, overlay customization (brushes, floating badges), the Validation Flow, and SDK version migration.
Laravel architecture patterns, routing/controllers, Eloquent ORM, service layers, queues, events, caching, and API resources for production apps. Use when building or reviewing Laravel apps — controllers, Eloquent, service layers, queues, or API resources.
Activate when the user analyzes a complex problem from only a single perspective (e.g. purely financial / purely technical / purely market), or lacks an analytical framework for a cross-domain problem. Typical triggers: the user says "I only understand XX" or "this problem seems to be an XX problem" while clearly relying on a single discipline's explanation. Not applicable when the user already has a mature multidimensional analytical framework and is using it effectively, nor to simple purely execution-level tasks.
Analyze problems through Lei Jun's lens of extreme cost-effectiveness, the "tuyere" theory and internet thinking. Use when the user explicitly asks for this person's perspective, mentions "leijun-perspective", or wants to apply his core methods to analyze a problem; do not trigger it automatically for general questions.
Build and maintain a data inventory (mapa de dados) of all personal data processing activities in the project. Use when the user asks “mapa de dados,” “data mapping LGPD,” “que dados pessoais coletamos,” or “inventário de tratamento,” or as part of an LGPD audit pipeline. Produces `.lgpd/data-map.md`. Especially important for legacy systems where the inventory is reverse-engineered from code (Prisma schemas, API endpoints, third-party SDKs, log statements)
Verify scholarly references and claim-citation support for Light stage 10. Use when auditing a manuscript, claim map, bibliography, DOI/arXiv/PMID/ISBN/URL, BibTeX/CSL, citekeys, chimeric or fabricated citations, retraction/correction alerts, or preparing a canonical citation registry for typesetting. Builds provenance-preserving inventories, confirms metadata with independent authoritative sources, distinguishes CONFIRMED/CONFIRMED-MISSING/UNAVAILABLE/UNRESOLVED, records Crossref update direction, and emits the citation gate plus delivery artifacts.
Light on-demand engineering skill · frontend design: turn a vague "make it look good" into runnable React/Tailwind/shadcn code + design decision notes — with a visual signature element, self-consistent style (consistent design tokens), scenario fit (academic poster/data dashboard/admin back office/mobile/marketing landing each have different information density), anti-"obviously AI" (purple-blue gradients/Inter/16px radii/giant hero/centered stacking/emoji headings), and visual accessibility (WCAG 2.2). When to use: competition entries/project demos/research system interfaces that need a web page or app UI / reworking an existing frontend / choosing a design system, palette or font / building dashboard, admin, landing or mobile UI. Trigger words: frontend / interface / UI / web page / website / design a page / landing page / landing / dashboard / data dashboard / admin back office / admin / component / React / Vue / Tailwind / shadcn / palette / font / design system / design system / make it look good / frontend design / redesign the interface / redesign. Core discipline: never decide the design direction/tech stack/palette/font for the user — these are decision points; give recommendations + reasons + alternatives and stop to ask the user via AskUserQuestion. Zero local design library: look up inspiration/trends/stack versions online (Awwwards / `npm view` / the WCAG site), mark last_checked, mark unknown if not found, never rely on a stale local library. Deliver runnable code, not empty talk; reuse `_shared/visual_qa` to judge contrast (do not reinvent the WCAG math). Not a DAG node (on demand: produces no findings, does not block the main line).
Light research integrity and ethics lifecycle always-on red-line gate: at research triage, approval, collection, change, analysis, submission, release, and post-publication, check risks including academic misconduct, data fabrication, statistical consistency, exaggerated conclusions, hallucinated or retracted citations, self-plagiarism, privacy, copyright, authorship and AI disclosure, software copyright and patent ownership, and paper-mill rewriting, turning "don't fabricate, don't exaggerate" from verbal advice into a machine-checkable, blockable, machine-readable gate aggregateable by the master run_checkpoint (produces light.findings.v1, Critical fail → exit 1). When to use: before submission or camera-ready, before releasing data or code, before filing software copyright, and when designing human-subjects or animal experiments — these are hard gates; plus background integrity scanning for any writing, analysis, or citation task. Trigger words: ethics / integrity / academic misconduct / fabrication / p-hacking / exaggeration / SOTA / retraction / plagiarism check / copying / privacy / IRB / informed consent / authorship / AI disclosure / paper mill / software copyright ownership. Core discipline: signal is not a verdict, and AI cannot self-assess → always "machine gate + human review"; when nothing is found, write "pending verification/UNRESOLVED", never fabricate; verify online throughout, zero local knowledge base, zero paid keys.
Light research mainline step 5 · research plan and experiment design: break the idea cleared by idea-critique and the data feasibility into a truly executable, paper-ready, reproducible question/estimand, experiment matrix and preregistration package. When to use: the idea has passed review and needs to land / experiments, ablations, comparisons, sensitivity, generalization, robustness need designing / writing a research plan PROJECT_PLAN / locking the primary outcome, exclusion/stopping rules or preregistration / planning sample size, seeds and statistical power / computing an experiment compute budget / reproducing an existing paper / worrying that the baseline is watered down or the hypothesis cannot be falsified. Trigger words: research plan / experiment design / experiment matrix / hypothesis / control baseline / ablation / fair comparison / falsifiable / statistical power / how many seeds / reproduction / reproducibility / research plan / experiment design / reproducible. Core discipline: unfair comparison (watered-down baseline) / non-falsifiability = critical veto (spec §4.2, STAGE_GATES[5]=[fair_baseline,falsifiable]); an ablation that doesn't isolate / underpowered = warn; power is back-computed at the planning stage, not a post-hoc result check; lint is heuristic and bounded, and the final verdict on fairness/falsifiability still needs human/domain judgement — never claim "proved the comparison is absolutely fair".
Build evidence-bound journal or conference shortlists for Light stage 12. Use after typesetting delivers venue-handoff.json/PDF/compliance facts; when an author asks where to submit, journal selection, conference fit, scope or article-type matching, publication strategy, reach/match/safety tiers, transfer order, APC/OA/indexing/deadline constraints, or predatory/hijacked-journal risk. Produces a current-source candidate registry, fit/risk/unknown reports, and an unchosen decision packet; never recompiles the PDF, invents acceptance rates, condemns a venue from soft signals, chooses without a direct user choice or explicit delegation, or submits.
A line editor that reads prose sentence by sentence and shows which lines are weak — filtering, overwriting, flat rhythm, blurred clarity, voice slips — naming the craft principle behind each and pointing at the fix, without ever writing the replacement. It works pattern-first: it surfaces the writer's recurring line-level habits and flags the clearest instances, so they fix the class in their own voice. Use it after the story works, when the prose needs to earn its place line by line. It marks and explains; it never rewrites the author's sentences.
Validates Stories/Tasks with GO/NO-GO verdict, Readiness Score (1-10), Penalty Points, and Anti-Hallucination verification. Auto-fixes to reach 0 points, delegates to ln-002 for docs. Use when reviewing Stories before execution or when user requests validation.
Fill in Word document form templates (.docx) with user-provided data. Reads a template containing tables with label→value cell pairs, detects all fillable fields, and outputs a completed document. Handles CJK/Latin mixed text with proper font switching. Use this skill when the user wants to fill in a form template, complete an application form, populate a Word table form, or automate document filling. Also trigger when the user mentions "填表", "填写表格", "fill form", "fill template", "表格填写", "申请表", "登记表", or has a .docx template with blank fields to fill.
Keep the durable record that outlives a session — credential provenance, access inventory, artifacts left on target for cleanup, findings with evidence, and a dead-end log — so work spanning days or analysts does not restart or contradict itself. Use when an engagement or investigation runs longer than one sitting, when resuming work after a gap, when handing off to another analyst, before a context window rolls over, or when you cannot say where a credential or access came from.
Use when optimizing content for AI search engines and answer platforms (ChatGPT, Perplexity, Gemini, Google AI Overviews). Covers GEO: crawl controls (robots/WAF/llms.txt), answer-ready content and entity pages, citation strategy, and measurement (query bank, share of model). For building monitoring infrastructure, see project-aeo-monitoring-tools.
Explains the capabilities and limits, supported formats, and ecosystem (CLI / Python API / plugins / Docker / Azure integration) of MarkItDown, Microsoft’s “convert anything to Markdown” tool. Guides the LLM to choose the right path for requests to convert PDF/PPT/Word/Excel/images/audio/web pages/YouTube to Markdown: the markitdown-awesome overview, markitdown-cli command guidance, or the markitdown-ocr LLM Vision OCR path. Use this skill when users mention markitdown, microsoft/markitdown, converting Office documents to markdown, OCR for scans, or related alternatives (textract, unstructured, pandoc).
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.