Full index · agent skills
All agent skills - page 148 of 640
Every one of the 63,961 agent skills in the directory, ranked by adoption. Jump to a tool, or browse the agent skills hub.
- Creative Strategy EngineGenerate ad creative briefs, angles, hooks, and channel plans so growth teams can ship consistent campaigns faster across paid social and lifecycle touchpoints.529 installs
- Credential ScannerUse credential-scanner for development tasks529 installs
- Longbridge Peer ComparisonGenerate side-by-side valuation and growth comparison matrices for 2-5 tickers via the Longbridge CLI.529 installs
- Metabase Representation FormatGenerate or validate Metabase query and visualization representation formats when embedding analytics, syncing dashboards, or building BI integrations in agent workflows.529 installs
- React Flow ArchitectureDecide whether React Flow fits your product and how to structure node-based UI state before you implement a workflow or diagram editor.529 installs
- React Ui PatternsGenerate consistent, production-grade React component patterns and UI implementations quickly.529 installs
- Reddit MarketingRun authentic Reddit community marketing: subreddit research, value-first posts, AMA planning, and comment engagement without spam bans or brand damage.529 installs
- Add AudioAdd background music, sound effects, and spatial audio to their game project with a single agent command.528 installs
- Analyzing Command And Control CommunicationGuide your agent through structured analysis of command-and-control (C2) traffic and infrastructure when you are hardening or investigating compromised systems.528 installs
- Cpp TemplatesDesign type-safe generic C++ libraries, metaprogrammed containers, and compile-time abstractions for performance-critical backend or engine modules.528 installs
- Create Agent With Sanity ContextGive AI agents structured, schema-aware access to their Sanity CMS content instead of relying on embeddings alone.528 installs
- Create Technical SpikeTime-box research spikes so critical technical questions are answered before your agent or team commits to an implementation path.528 installs
- DcgIntercept or block rm -rf, force pushes, credential wipes, and other irreversible shell commands before an agent or developer executes them in local or CI environments.528 installs
- Kelly CriterionCompute Kelly-optimal, risk-adjusted position sizes from win rate and payoff ratio.528 installs
- Longbridge AlertSet, check, or remove a Longbridge price alert with an explicit two-step confirm.528 installs
- Mssqlmssql: agent skill for task automation in Claude Code workflows.528 installs
- Neo4j Cli Tools SkillOperate Neo4j CLI tools for schema inspection, Cypher execution, import/export, and local graph administration while building or debugging graph-backed application features.528 installs
- Release ManagerAutomatically prepare, version, and coordinate production releases from conventional commit history.528 installs
- Research AgentLet Claude autonomously gather market data, competitor analysis, technical documentation, and user insights before starting any new feature or product.528 installs
- Rust Sanitizers MiriRun Miri and LLVM sanitizers on unsafe or FFI-heavy Rust to catch UB, data races, and memory errors before merging low-level systems code.528 installs
- Analyzing Dns Logs For ExfiltrationRun structured DNS log reviews to spot tunneling, C2 beaconing, and exfiltration when you operate your own apps or small-team production DNS telemetry.527 installs
- Ax AuditRun accessibility audits on pages or components to find WCAG violations, keyboard traps, and contrast issues before public launch.527 installs
- CodeqlTriage, prioritize, dismiss, and resolve GitHub CodeQL code-scanning alerts using severity and CVSS rules.527 installs
- Create SpecificationGenerate a machine-readable Markdown spec under /spec/ so your coding agent can implement features without ambiguous requirements.527 installs
- Creative MechanicsDesign reusable creative campaign mechanics—hooks, loops, and narrative frames—that improve engagement and conversion across paid and organic channels.527 installs
- Enterprise Architecture PatternsChoose enterprise architecture patterns—layering, boundaries, messaging, and scalability—before building large multi-team SaaS or API platforms.527 installs
- GlassmorphismDrop frosted-glass overlays and floating controls into a React/Tailwind UI without guessing blur, opacity, and border tokens.527 installs
- Grpc Service DevelopmentGenerate complete, production-ready gRPC service definitions, Protocol Buffer schemas, and streaming implementations for microservices.527 installs
- Here Be GitNavigate git history, branches, and PR hygiene safely so agents inspect repos, prepare commits, and avoid destructive version-control mistakes.527 installs
- Intent ModelingPause before executing author instructions and separate literal wording from intended outcome so the agent aligns with what the user actually needs.527 installs
- Longbridge DcaCreate, modify, pause, resume, or stop Longbridge recurring investment plans with safe two-step confirmation.527 installs
- Longbridge FxQuery FX cross-rates and convert multi-currency portfolio holdings without login527 installs
- Qa ReportProduce structured QA reports summarizing test runs, regressions, repro steps, severity, and release readiness after manual or automated verification cycles.527 installs
- Sanity Live Cache ComponentsUse sanity-live-cache-components for development tasks527 installs
- Apktoolsecurity-audit526 installs
- Commit Message StorytellerTurn messy git diffs into conventional, reviewer-friendly commit messages with a clear narrative body.526 installs
- Concurrency DebuggingDiagnose race conditions, deadlocks, lock contention, and nondeterministic failures in multithreaded C/C++/Rust services before release.526 installs
- Configuring Connected AppsCorrectly configure Salesforce Connected Apps and External Client Apps with proper OAuth, JWT bearer, scopes, and security settings.526 installs
- Create AgentsmdGenerate a root-level AGENTS.md that gives Claude Code, Cursor, and other agents predictable setup, commands, and repo conventions without bloating the human README.526 installs
- Create ReadmeGenerate a polished, GitHub-ready README.md after reviewing the whole repo so newcomers and install instructions are clear.526 installs
- Dotnet Local ToolsConfigure and use dotnet local tools manifests so teams pin CLI utilities—formatters, generators, analyzers—without polluting global SDK installs across repos.526 installs
- Fal Models CatalogBrowse and compare fal.ai model endpoints, capabilities, and pricing before choosing image, video, or audio generators for an agent workflow or product feature.526 installs
- Gap AnalysisCompare current capabilities, competitor offerings, and target requirements to define MVP scope, priorities, and missing features before build commitments.526 installs
- Linkedin Profile OptimizerRewrite LinkedIn headline, about, experience, and featured sections so founders and sellers sound credible to buyers searching LinkedIn and discovering profiles via AI summaries.526 installs
- Longbridge Etf FlowTracks US sector ETF inflows and outflows to read sector rotation and risk appetite.526 installs
- Longbridge Watchlist AdminMutate Longbridge watchlist groups, symbols, renames, or deletes via preview-then-confirm flow after login and explicit user command.526 installs
- Modals DialogsAnimate modal open, close, and focus trap flows for confirmations, forms, and drawers so overlays feel grounded and dismiss gestures stay predictable.526 installs
- Network DebuggingDiagnose connectivity, DNS, TLS, latency, and API failures using curl, tcpdump, traceroute, and browser or proxy tooling.526 installs
- Pr ReviewRun a full, framework-aware Angular pull request review with package-specific rules, commit-message quality, and backwards-compatibility checks.526 installs
- Retail ExpertFrame retail product requirements—catalog, checkout, inventory, promotions, and store ops—before building so scope matches real merchandising and omnichannel constraints.526 installs
- Sql Injection PreventionHarden database-backed APIs and forms against SQL injection during security review before production release.526 installs
- Supabase Audit RpcDiscover and test PostgreSQL functions exposed via Supabase RPC endpoints, auditing authentication and authorization controls to find vulnerabilities before shipping.526 installs
- Verification Before CompletionForce the agent to run fresh verification commands and cite evidence before saying tests pass, builds succeed, or work is complete.526 installs
- Web Performance AuditUse web-performance-audit for development tasks526 installs
- Animator TraditionalTranslate classical hand-drawn animation thinking—keyframes, arcs, spacing—into digital sprites, SVG sequences, or rigged character motion.525 installs
- AutoresearchAutomate literature and note synthesis inside Obsidian vaults—collect sources, summarize findings, and link insights before committing to a product direction.525 installs
- Btc Bottom ModelGet systematic, weighted Bitcoin cycle analysis with a 0-100 Market Heat Score and clear buy or sell signals instead of relying on gut feel or single indicators.525 installs
- Create Implementation PlanGenerate a deterministic, machine-readable implementation plan with phased atomic tasks before your agent or you execute a feature, refactor, or upgrade.525 installs
- Create LlmsBootstrap a spec-compliant root llms.txt so coding agents and LLM crawlers can discover your repo’s purpose, layout, and key entry docs.525 installs
- Create Tldr PageTurn upstream CLI or tool docs into a concise tldr-pages-style reference with real command examples.525 installs
- DependabotAuthor or fix `.github/dependabot.yml` with correct ecosystem keys, schedules, groups, and private registry blocks.525 installs
- Entrance AnimationsAnimate first appearance of cards, modals, list rows, and hero sections with staged fades, slides, or scales that orient users without delaying interaction or harming Largest Contentful Paint.525 installs
- Forex ListRetrieve and normalize tradable forex pair listings for macro research, cross-border revenue context, hedging analysis, and multi-currency financial modeling workflows.525 installs
- Github ReleaseClassify git history with Conventional Commits and freeform heuristics so release notes and ship decisions trust the diff over vague messages.525 installs
- Instagram Content BenchmarkBenchmark Instagram posts and Reels against themes to extract winning hooks, formats, and caption patterns for organic growth525 installs
- Karpathytalk CommunitySelf-host and interact with a markdown-based developer social network that is natively friendly to LLM agents.525 installs
- Marketing Demand AcquisitionCompute blended and per-channel customer acquisition cost from spend and signup counts for campaign reviews.525 installs
- PrototypeTurn early product ideas into clickable Adobe prototypes that test navigation, layout, and core flows before committing engineering capacity.525 installs
- Rest Api Design PatternsModel REST resources, versioning, pagination, errors, auth, and contracts for public or internal HTTP APIs consumed by web and mobile clients.525 installs
- Shared DaoKeep canonical story vocabulary in kb/ so agents do not rename magic systems, factions, or genres mid-draft.525 installs
- Twitter CliPost updates, reply to mentions, and run scheduled campaigns directly from the terminal or agent workflows without switching to the Twitter web app.525 installs
- Ux ResearcherRun UX research informed by animation principles to test whether motion helps comprehension, reduces errors, and matches user mental models before full build.525 installs
- DocxGenerate clean, structured Microsoft Word documents directly from Claude conversations or agent instructions.524 installs
- Excitement EnergyAdd celebratory bursts, springy feedback, and upbeat pacing to onboarding wins, achievements, and high-energy product moments without breaking usability.524 installs
- Hindsight Self HostedUse hindsight-self-hosted for development tasks524 installs
- Longbridge ConsensusShows the analyst consensus for a company, including EPS and target-price estimates, ratings and revision trends.524 installs
- Longbridge Sec FilingsLets a solo builder pull a stock's SEC 10-K, 10-Q, 8-K, proxy and Form 4 filings and get key risks, KPIs and insider activity without reading raw EDGAR documents.524 installs
- Longbridge Security ListGives a solo builder a lookup of US overnight-tradable securities and a Hong Kong broker ID to name mapping for quick reference.524 installs
- Memory ResearchRun structured research sessions where findings are captured into durable agent memory for synthesis, citation, and follow-up questions across days or weeks.524 installs
- Multi Person TaskSplits one task across several people or agents so a solo builder can coordinate parallel work without losing track of owners.524 installs
- Payment Gateway IntegrationIntegrate Stripe, PayPal, or similar gateways with checkout flows, webhooks, idempotent charges, refunds, subscriptions, and PCI-aware server-side handling for monetized products.524 installs
- Pay With AppHelps with ai & agent building tasks.524 installs
- Perplexity SearchLet their coding agent retrieve fresh, cited web research from Perplexity directly inside Claude or Cursor sessions.524 installs
- PushHelps with ai & agent building tasks.524 installs
- RevenuecatPulls MRR, churn, customer and subscription data from RevenueCat so a solo app builder can check revenue without the dashboard.524 installs
- Spatial ThinkingApply spatial-thinking animation principles when designing motion, layout transitions, or interactive UI so movement feels intentional, readable, and on-brand.524 installs
- Subagent Driven DevelopmentExecute an implementation plan by dispatching a fresh subagent per task with spec then quality review.524 installs
- Write PlanTurn an ambiguous or high-risk change request into a phased, verifiable implementation plan that another agent can execute.524 installs
- Anomaly DetectionDetect unusual metrics, logs, or time-series spikes in production data using statistical and ML methods to catch incidents and fraud early.523 installs
- Arc MasteryAnimate drags, reveals, and cursor-driven elements along curved paths instead of stiff linear slides for more natural interface motion.523 installs
- Building Sf IntegrationsCorrectly configure Salesforce Named Credentials, External Services, REST/SOAP callouts, Platform Events, and Change Data Capture using battle-tested architecture523 installs
- Customer SupportDesign AI-assisted support workflows—chatbots, ticketing, sentiment, and omnichannel playbooks—for developers who own post-launch customer experience.523 installs
- Design MdDraft structured design markdown for features, APIs, or systems before implementation so scope, tradeoffs, and rollout plans are agreed and traceable.523 installs
- Docker DeploymentContainerize a Node.js service with multi-stage images, non-root users, health checks, and Compose so you can ship the same build locally and in prod.523 installs
- Encore Go Getting StartedBootstrap an Encore.go backend with services, infra-as-code primitives, local dev workflows, and deployable API foundations for a new SaaS or internal tool.523 installs
- Firecrawl ScraperLet their coding agent scrape websites, crawl pages, extract structured data, capture screenshots, or parse PDFs using the Firecrawl API.523 installs
- GgsqlUse ggsql for development tasks523 installs
- Linker ScriptsAuthor and debug custom linker scripts that place code, data, and BSS into flash, RAM, and peripheral regions for bare-metal, RTOS, and kernel-linked binaries.523 installs
- Research GrantsQuickly surface relevant grant opportunities, eligibility criteria, and application requirements using Claude.523 installs