NZETA.AI PICKS · 2026
One directory for the best AI agents and the practical skills that make them useful in 2026.
View RankingsOVERALL RANKING
Ranked by agent capability, practical value, product influence, and ecosystem maturity. Updated 2026-06-15.
An all-purpose agent for web research, files, analysis, and multi-step tasks
A strong reasoning and writing assistant for complex knowledge work
A multimodal assistant with deep Google ecosystem integration
A cloud coding agent for parallel software engineering tasks
A terminal-first coding agent that understands entire codebases
An enterprise platform for building and governing Microsoft-connected agents
CRM-native agents for sales, service, and business operations
Business agents connected to thousands of popular applications
A flexible, self-hostable platform for AI agents and workflows
A customer service agent for resolving support questions across channels
CATEGORY RANKINGS
Switch categories without losing the compact ranking view. Each category keeps the list short and comparison-ready.
All-purpose agents for everyday work, content, analysis, and multi-step tasks.
An all-purpose agent for web research, files, analysis, and multi-step tasks
A strong reasoning and writing assistant for complex knowledge work
A multimodal assistant with deep Google ecosystem integration
A productivity assistant connected to the web and Microsoft 365
A general assistant with real-time information access
An autonomous agent focused on completing end-to-end tasks
A source-first research agent for fast, cited answers
A long-context assistant for research and document analysis
An AI workspace for research, content, and generated deliverables
A deep research agent for professional reports and analysis
Agents built for web research, source discovery, synthesis, and long-form analysis.
A source-first research agent for fast, cited answers
A multimodal assistant with deep Google ecosystem integration
An all-purpose agent for web research, files, analysis, and multi-step tasks
A strong reasoning and writing assistant for complex knowledge work
A deep research agent for professional reports and analysis
An AI workspace for research, content, and generated deliverables
An autonomous agent focused on completing end-to-end tasks
A long-context assistant for research and document analysis
A productivity assistant connected to the web and Microsoft 365
A general assistant with real-time information access
Agents that understand codebases, edit files, run tests, and complete engineering tasks.
A cloud coding agent for parallel software engineering tasks
A terminal-first coding agent that understands entire codebases
An AI coding platform embedded across GitHub, IDE, and CLI workflows
An AI-native code editor with interactive and background agents
An autonomous software engineering agent for end-to-end development tasks
A build-and-deploy agent for turning ideas into working applications
An agentic development environment for codebase-wide changes
An asynchronous cloud coding agent connected to GitHub repositories
A development agent for coding, modernization, and AWS workflows
An agentic coding tool for complex codebase tasks
Agents that connect apps, data, tools, and approvals into repeatable workflows.
Business agents connected to thousands of popular applications
A flexible, self-hostable platform for AI agents and workflows
No-code AI employees for common business operations
A low-code platform for building multi-agent workforces
A visual automation platform combining AI, web data, and business tools
Agent-driven decisions inside visual Make automation scenarios
An open-source platform for agents, knowledge, and generative AI workflows
AI agents for web research, enrichment, sales, and operations
Human-in-the-loop AI workflows with app integrations and approvals
A framework and platform for orchestrating multi-agent teams
Agent platforms focused on governance, enterprise data, permissions, and deployment.
An enterprise platform for building and governing Microsoft-connected agents
CRM-native agents for sales, service, and business operations
Enterprise search and agent experiences connected to company data
Agents designed for enterprise service and workflow operations
Governed agent orchestration for complex enterprise workflows
Business agents embedded in Oracle Fusion Applications
AWS-native agents connected to models, knowledge, and APIs
A platform combining AI agents, robots, and human workflows
Business agents operating across SAP data and processes
Agents for governed HR and finance workflows
Agents for resolving customer questions, handling tickets, and improving service operations.
A customer service agent for resolving support questions across channels
A customer experience agent that can complete transactions for brands
A generative customer support agent for chat, email, and voice
A multilingual customer service automation platform
Support agents for ticket resolution, routing, and agent assistance
Enterprise voice agents for high-volume customer service calls
A developer platform for real-time voice agents and phone automation
A programmable platform for inbound and outbound phone agents
Agents designed for enterprise service and workflow operations
CRM-native agents for sales, service, and business operations
Agents for prospect research, outreach, qualification, follow-up, and CRM workflows.
A digital sales development agent for prospecting and outreach
An AI sales development agent for B2B website visitors
AI agents for web research, enrichment, sales, and operations
No-code AI employees for common business operations
A low-code platform for building multi-agent workforces
Business agents connected to thousands of popular applications
CRM-native agents for sales, service, and business operations
An enterprise platform for building and governing Microsoft-connected agents
A customer experience agent that can complete transactions for brands
A programmable platform for inbound and outbound phone agents
Agents that handle inbound and outbound calls, appointments, sales, and phone support.
A developer platform for real-time voice agents and phone automation
Enterprise voice agents for high-volume customer service calls
A programmable platform for inbound and outbound phone agents
A generative customer support agent for chat, email, and voice
A multilingual customer service automation platform
A customer experience agent that can complete transactions for brands
A customer service agent for resolving support questions across channels
Support agents for ticket resolution, routing, and agent assistance
A digital sales development agent for prospecting and outreach
An AI sales development agent for B2B website visitors
COMPARE AI AGENTS
Pricing and product capabilities change frequently. Verify details on the official website before purchasing.
| Product | Category | Pricing | Free | API | Self-hosted | Integrations | Best for | Pros | Limitations | Official |
|---|---|---|---|---|---|---|---|---|---|---|
| CAOpenAIChatGPT agent | General Assistants | Free tier + paid plans | Yes | Yes | No | Web, Files, Productivity tools | General productivity, research, writing, and multi-step knowledge work | Broad task coverage; Easy to start | Usage limits vary by plan; High-impact actions still need review | Visit ↗ |
| CAnthropicClaude | General Assistants | Free tier + paid plans | Yes | Yes | No | Web, Files, Productivity tools | General productivity, research, writing, and multi-step knowledge work | Broad task coverage; Easy to start | Usage limits vary by plan; High-impact actions still need review | Visit ↗ |
| GGoogleGemini | General Assistants | Free tier + paid plans | Yes | Yes | No | Web, Files, Productivity tools | General productivity, research, writing, and multi-step knowledge work | Broad task coverage; Easy to start | Usage limits vary by plan; High-impact actions still need review | Visit ↗ |
| OCOpenAIOpenAI Codex | Coding | Subscription + usage | No | Yes | No | Git repositories, IDEs, Terminal | Software development, codebase changes, testing, and engineering tasks | Strong developer workflow; Handles multi-file tasks | Performance varies by repository; Generated changes require review | Visit ↗ |
| CCAnthropicClaude Code | Coding | Subscription + usage | No | Yes | No | Git repositories, IDEs, Terminal | Software development, codebase changes, testing, and engineering tasks | Strong developer workflow; Handles multi-file tasks | Performance varies by repository; Generated changes require review | Visit ↗ |
| MCMicrosoftMicrosoft Copilot Studio | Enterprise Platforms | Enterprise quote | No | Yes | No | Enterprise apps, Identity systems, Company data | Governed enterprise agents connected to company data and business systems | Enterprise controls; Deep platform integration | Longer implementation cycle; Pricing often requires a sales quote | Visit ↗ |
| SASalesforceSalesforce Agentforce | Enterprise Platforms | Enterprise quote | No | Yes | No | Enterprise apps, Identity systems, Company data | Governed enterprise agents connected to company data and business systems | Enterprise controls; Deep platform integration | Longer implementation cycle; Pricing often requires a sales quote | Visit ↗ |
| ZAZapierZapier Agents | Workflow Automation | Subscription + usage | Yes | Yes | No | SaaS apps, APIs, Databases | Connecting business apps and automating repeatable multi-step workflows | Fast workflow deployment; Reusable automations | Complex workflows need maintenance; Usage costs can scale quickly | Visit ↗ |
| Nn8nn8n | Workflow Automation | Subscription + usage | Yes | Yes | Yes | SaaS apps, APIs, Databases | Connecting business apps and automating repeatable multi-step workflows | Fast workflow deployment; Reusable automations | Complex workflows need maintenance; Usage costs can scale quickly | Visit ↗ |
| IFIntercomIntercom Fin | Customer Support | Paid plans | No | Yes | No | Help desks, CRM, Knowledge bases | Customer support automation, ticket resolution, and service operations | Reduces repetitive tickets; Supports human escalation | Quality depends on support content; Complex cases need human agents | Visit ↗ |
| MCMicrosoftMicrosoft Copilot | General Assistants | Free tier + paid plans | Yes | No | No | Web, Files, Productivity tools | General productivity, research, writing, and multi-step knowledge work | Broad task coverage; Easy to start | Usage limits vary by plan; High-impact actions still need review | Visit ↗ |
| GxAIGrok | General Assistants | Free tier + paid plans | Yes | Yes | No | Web, Files, Productivity tools | General productivity, research, writing, and multi-step knowledge work | Broad task coverage; Easy to start | Usage limits vary by plan; High-impact actions still need review | Visit ↗ |
| MManusManus | General Assistants | Paid plans | No | No | No | Web, Files, Productivity tools | General productivity, research, writing, and multi-step knowledge work | Broad task coverage; Easy to start | Usage limits vary by plan; High-impact actions still need review | Visit ↗ |
| PPerplexityPerplexity | General Assistants | Free tier + paid plans | Yes | Yes | No | Web, Files, Productivity tools | General productivity, research, writing, and multi-step knowledge work | Broad task coverage; Easy to start | Usage limits vary by plan; High-impact actions still need review | Visit ↗ |
| KMoonshot AIKimi | General Assistants | Free tier + paid plans | Yes | Yes | No | Web, Files, Productivity tools | General productivity, research, writing, and multi-step knowledge work | Broad task coverage; Easy to start | Usage limits vary by plan; High-impact actions still need review | Visit ↗ |
| GMainFuncGenspark | General Assistants | Paid plans | No | No | No | Web, Files, Productivity tools | General productivity, research, writing, and multi-step knowledge work | Broad task coverage; Easy to start | Usage limits vary by plan; High-impact actions still need review | Visit ↗ |
| YAYou.comYou.com ARI | General Assistants | Paid plans | No | No | No | Web, Files, Productivity tools | General productivity, research, writing, and multi-step knowledge work | Broad task coverage; Easy to start | Usage limits vary by plan; High-impact actions still need review | Visit ↗ |
| GCGitHubGitHub Copilot | Coding | Free tier + paid plans | Yes | No | No | Git repositories, IDEs, Terminal | Software development, codebase changes, testing, and engineering tasks | Strong developer workflow; Handles multi-file tasks | Performance varies by repository; Generated changes require review | Visit ↗ |
| CAnysphereCursor | Coding | Free tier + paid plans | Yes | No | No | Git repositories, IDEs, Terminal | Software development, codebase changes, testing, and engineering tasks | Strong developer workflow; Handles multi-file tasks | Performance varies by repository; Generated changes require review | Visit ↗ |
| DCognitionDevin | Coding | Paid plans | No | No | No | Git repositories, IDEs, Terminal | Software development, codebase changes, testing, and engineering tasks | Strong developer workflow; Handles multi-file tasks | Performance varies by repository; Generated changes require review | Visit ↗ |
| RAReplitReplit Agent | Coding | Free tier + paid plans | Yes | No | No | Git repositories, IDEs, Terminal | Software development, codebase changes, testing, and engineering tasks | Strong developer workflow; Handles multi-file tasks | Performance varies by repository; Generated changes require review | Visit ↗ |
| WWindsurfWindsurf | Coding | Free tier + paid plans | Yes | No | No | Git repositories, IDEs, Terminal | Software development, codebase changes, testing, and engineering tasks | Strong developer workflow; Handles multi-file tasks | Performance varies by repository; Generated changes require review | Visit ↗ |
| JGoogleJules | Coding | Free tier + paid plans | Yes | No | No | Git repositories, IDEs, Terminal | Software development, codebase changes, testing, and engineering tasks | Strong developer workflow; Handles multi-file tasks | Performance varies by repository; Generated changes require review | Visit ↗ |
| AQAWSAmazon Q Developer | Coding | Free tier + paid plans | Yes | Yes | No | Git repositories, IDEs, Terminal | Software development, codebase changes, testing, and engineering tasks | Strong developer workflow; Handles multi-file tasks | Performance varies by repository; Generated changes require review | Visit ↗ |
| ASourcegraphAmp | Coding | Paid plans | No | No | No | Git repositories, IDEs, Terminal | Software development, codebase changes, testing, and engineering tasks | Strong developer workflow; Handles multi-file tasks | Performance varies by repository; Generated changes require review | Visit ↗ |
| LLindyLindy | Workflow Automation | Subscription + usage | Yes | Yes | No | SaaS apps, APIs, Databases | Connecting business apps and automating repeatable multi-step workflows | Fast workflow deployment; Reusable automations | Complex workflows need maintenance; Usage costs can scale quickly | Visit ↗ |
| RARelevance AIRelevance AI | Workflow Automation | Subscription + usage | Yes | Yes | No | SaaS apps, APIs, Databases | Connecting business apps and automating repeatable multi-step workflows | Fast workflow deployment; Reusable automations | Complex workflows need maintenance; Usage costs can scale quickly | Visit ↗ |
| GGumloopGumloop | Workflow Automation | Subscription + usage | Yes | Yes | No | SaaS apps, APIs, Databases | Connecting business apps and automating repeatable multi-step workflows | Fast workflow deployment; Reusable automations | Complex workflows need maintenance; Usage costs can scale quickly | Visit ↗ |
| MAMakeMake AI Agents | Workflow Automation | Subscription + usage | Yes | Yes | No | SaaS apps, APIs, Databases | Connecting business apps and automating repeatable multi-step workflows | Fast workflow deployment; Reusable automations | Complex workflows need maintenance; Usage costs can scale quickly | Visit ↗ |
| DDifyDify | Workflow Automation | Free tier + paid plans | Yes | Yes | Yes | SaaS apps, APIs, Databases | Connecting business apps and automating repeatable multi-step workflows | Fast workflow deployment; Reusable automations | Complex workflows need maintenance; Usage costs can scale quickly | Visit ↗ |
| BBardeenBardeen | Workflow Automation | Free tier + paid plans | Yes | Yes | No | SaaS apps, APIs, Databases | Connecting business apps and automating repeatable multi-step workflows | Fast workflow deployment; Reusable automations | Complex workflows need maintenance; Usage costs can scale quickly | Visit ↗ |
| RRelayRelay.app | Workflow Automation | Free tier + paid plans | Yes | Yes | No | SaaS apps, APIs, Databases | Connecting business apps and automating repeatable multi-step workflows | Fast workflow deployment; Reusable automations | Complex workflows need maintenance; Usage costs can scale quickly | Visit ↗ |
| CCrewAICrewAI | Workflow Automation | Free tier + paid plans | Yes | Yes | Yes | SaaS apps, APIs, Databases | Connecting business apps and automating repeatable multi-step workflows | Fast workflow deployment; Reusable automations | Complex workflows need maintenance; Usage costs can scale quickly | Visit ↗ |
| GAGoogle CloudGoogle Agentspace | Enterprise Platforms | Enterprise quote | No | Yes | No | Enterprise apps, Identity systems, Company data | Governed enterprise agents connected to company data and business systems | Enterprise controls; Deep platform integration | Longer implementation cycle; Pricing often requires a sales quote | Visit ↗ |
| SAServiceNowServiceNow AI Agents | Enterprise Platforms | Enterprise quote | No | Yes | No | Enterprise apps, Identity systems, Company data | Governed enterprise agents connected to company data and business systems | Enterprise controls; Deep platform integration | Longer implementation cycle; Pricing often requires a sales quote | Visit ↗ |
| IWIBMIBM watsonx Orchestrate | Enterprise Platforms | Enterprise quote | Yes | Yes | No | Enterprise apps, Identity systems, Company data | Governed enterprise agents connected to company data and business systems | Enterprise controls; Deep platform integration | Longer implementation cycle; Pricing often requires a sales quote | Visit ↗ |
| OAOracleOracle AI Agent Studio | Enterprise Platforms | Enterprise quote | No | Yes | No | Enterprise apps, Identity systems, Company data | Governed enterprise agents connected to company data and business systems | Enterprise controls; Deep platform integration | Longer implementation cycle; Pricing often requires a sales quote | Visit ↗ |
| ABAWSAmazon Bedrock Agents | Enterprise Platforms | Enterprise quote | No | Yes | No | Enterprise apps, Identity systems, Company data | Governed enterprise agents connected to company data and business systems | Enterprise controls; Deep platform integration | Longer implementation cycle; Pricing often requires a sales quote | Visit ↗ |
| UAUiPathUiPath Agentic Automation | Enterprise Platforms | Enterprise quote | No | Yes | No | Enterprise apps, Identity systems, Company data | Governed enterprise agents connected to company data and business systems | Enterprise controls; Deep platform integration | Longer implementation cycle; Pricing often requires a sales quote | Visit ↗ |
| SJSAPSAP Joule Agents | Enterprise Platforms | Enterprise quote | No | Yes | No | Enterprise apps, Identity systems, Company data | Governed enterprise agents connected to company data and business systems | Enterprise controls; Deep platform integration | Longer implementation cycle; Pricing often requires a sales quote | Visit ↗ |
| WAWorkdayWorkday AI Agents | Enterprise Platforms | Enterprise quote | No | Yes | No | Enterprise apps, Identity systems, Company data | Governed enterprise agents connected to company data and business systems | Enterprise controls; Deep platform integration | Longer implementation cycle; Pricing often requires a sales quote | Visit ↗ |
| SSierraSierra | Customer Support | Paid plans | No | Yes | No | Help desks, CRM, Knowledge bases | Customer support automation, ticket resolution, and service operations | Reduces repetitive tickets; Supports human escalation | Quality depends on support content; Complex cases need human agents | Visit ↗ |
| DDecagonDecagon | Customer Support | Paid plans | No | Yes | No | Help desks, CRM, Knowledge bases | Customer support automation, ticket resolution, and service operations | Reduces repetitive tickets; Supports human escalation | Quality depends on support content; Complex cases need human agents | Visit ↗ |
| AAdaAda | Customer Support | Paid plans | No | Yes | No | Help desks, CRM, Knowledge bases | Customer support automation, ticket resolution, and service operations | Reduces repetitive tickets; Supports human escalation | Quality depends on support content; Complex cases need human agents | Visit ↗ |
| FForethoughtForethought | Customer Support | Paid plans | No | Yes | No | Help desks, CRM, Knowledge bases | Customer support automation, ticket resolution, and service operations | Reduces repetitive tickets; Supports human escalation | Quality depends on support content; Complex cases need human agents | Visit ↗ |
| PPolyAIPolyAI | Customer Support | Paid plans | No | Yes | No | Help desks, CRM, Knowledge bases | Customer support automation, ticket resolution, and service operations | Reduces repetitive tickets; Supports human escalation | Quality depends on support content; Complex cases need human agents | Visit ↗ |
| RARetell AIRetell AI | Customer Support | Subscription + usage | No | Yes | No | Help desks, CRM, Knowledge bases | Customer support automation, ticket resolution, and service operations | Reduces repetitive tickets; Supports human escalation | Quality depends on support content; Complex cases need human agents | Visit ↗ |
| BABland AIBland AI | Customer Support | Subscription + usage | No | Yes | No | Help desks, CRM, Knowledge bases | Customer support automation, ticket resolution, and service operations | Reduces repetitive tickets; Supports human escalation | Quality depends on support content; Complex cases need human agents | Visit ↗ |
| 1A11x11x Alice | Sales | Paid plans | No | No | No | CRM, Email, Sales tools | Prospect research, outreach, qualification, follow-up, and CRM updates | Automates repetitive sales work; Improves follow-up consistency | Requires careful messaging controls; Deliverability and data quality vary | Visit ↗ |
| QPQualifiedQualified Piper | Sales | Paid plans | No | Yes | No | CRM, Email, Sales tools | Prospect research, outreach, qualification, follow-up, and CRM updates | Automates repetitive sales work; Improves follow-up consistency | Requires careful messaging controls; Deliverability and data quality vary | Visit ↗ |
TOP SKILLS
Top Skills are ranked by practical adoption value, composability, risk reduction, and how often they unlock repeatable workflows for modern AI agents and coding agents.
Break broad goals into ordered, checkable steps.
Call APIs, tools, and functions when natural language is not enough.
Navigate websites, click, type, compare, and submit web tasks.
Find and cite relevant information across docs, websites, and knowledge bases.
Understand project structure, dependencies, and implementation paths.
Read, summarize, compare, and transform uploaded files.
Clean, calculate, chart, and explain structured data.
Pause for review before sensitive or irreversible actions.
Use MCP servers and connectors to access many tools without bloating prompts.
Constrain agent behavior with permissions, policies, and risk checks.
SKILL CATEGORIES
Skills are grouped by how teams actually adopt agents: core capability, daily work, and technical production workflows.
Foundational skills that make agents useful across research, operations, coding, and enterprise workflows.
Break broad goals into ordered, checkable steps.
Call APIs, tools, and functions when natural language is not enough.
Navigate websites, click, type, compare, and submit web tasks.
Find and cite relevant information across docs, websites, and knowledge bases.
Maintain useful context across long tasks and repeat interactions.
Read, summarize, compare, and transform uploaded files.
Clean, calculate, chart, and explain structured data.
Pause for review before sensitive or irreversible actions.
Constrain agent behavior with permissions, policies, and risk checks.
Coordinate specialist agents around a shared outcome.
Skills that help agents complete everyday business tasks with files, documents, meetings, and knowledge bases.
Create polished docs from prompts, notes, data, or templates.
Turn research, outlines, or data into slide decks.
Summarize inboxes, draft replies, and coordinate schedules.
Research accounts, update CRM, draft outreach, and track follow-up.
Resolve routine tickets using knowledge, account context, and escalation rules.
Extract decisions, action items, owners, and next steps from meetings.
Search, compare, verify, and summarize web sources.
Complete repetitive internal forms and operational workflows.
Turn one source into posts, emails, landing pages, and summaries.
Find stale articles, draft updates, and connect support knowledge.
Skills for developers and operators who need agents to work with code, data, APIs, systems, and production workflows.
Understand project structure, dependencies, and implementation paths.
Make scoped code changes that follow local project patterns.
Create tests, reproduce failures, inspect logs, and propose fixes.
Run commands, scripts, data jobs, and local verification safely.
Connect services with authenticated calls and structured payloads.
Use MCP servers and connectors to access many tools without bloating prompts.
Route tasks through steps, agents, approvals, and tools.
Inspect, transform, validate, and move data between systems.
Measure agent quality, failures, costs, and repeatability.
Move agents from prototype to running cloud services.
Installable Claude Code skills and skill packs that extend coding agents with repeatable development, design, SEO, browser, diagram, video, and connector workflows.
A structured development methodology skill pack for disciplined agentic coding.
A frontend design intelligence skill for UI systems, palettes, typography, and UX rules.
A Claude Code SEO skill pack for audits, search research, and optimization workflows.
A browser automation skill for letting coding agents inspect and operate web pages.
A meta-skill for creating, packaging, and improving other skills.
A diagram generation skill for architecture, flows, and visual explanations.
A video composition skill for creating scripted, code-driven videos.
A rapid prototyping skill for rich HTML artifacts and interactive demos.
A connector skill for using MCP servers from Claude Code without custom setup.
A source-grounded research skill for citation-aware answers and synthesis.
COMPARE TOP SKILLS
Use this table to compare adoption, maturity, risk, inputs, outputs, and source evidence for each skill.
| Skill | Category | Adoption | Maturity | Risk | Best for | Inputs | Outputs | Why it matters | Evidence | Source |
|---|---|---|---|---|---|---|---|---|---|---|
| PTCore SkillsPlanning & Task Decomposition | Core Skills | Very high | High | Medium | Complex multi-step work, operations playbooks, research projects, and implementation plans. | Goal, Constraints, Success criteria | Task plan, Milestones, Dependencies | Good planning is the difference between a chatbot that answers and an agent that can finish work. | OpenAI describes agents as systems that plan, call tools, collaborate, and keep state for multi-step work. | Source ↗ |
| TUCore SkillsTool Use & Function Calling | Core Skills | Very high | High | Medium | Booking, CRM updates, calculations, retrieval, payments, operations, and app workflows. | Tool descriptions, Schemas, Permissions | Structured tool calls, Tool results, Action logs | Tool use turns agent reasoning into real action while preserving a structured interface for control. | Claude docs describe tool use as letting Claude call defined or provider tools based on the request and tool description. | Source ↗ |
| BOCore SkillsBrowser Operation | Core Skills | High | Medium-high | High | Web research, form filling, procurement, travel, lead enrichment, and repetitive web operations. | URL, Goal, Login/approval rules | Completed web task, Screenshots, Action trace | Many business tasks still live in web UIs, so browser skill is often the fastest path from intent to completed action. | OpenAI says ChatGPT agent can navigate websites, fill out forms, and work with uploaded files while keeping users in control. | Source ↗ |
| RKCore SkillsRetrieval & Knowledge Search | Core Skills | Very high | High | Medium | Enterprise Q&A, policy lookup, market research, support answers, and compliance review. | Knowledge sources, Query, Citation rules | Grounded answer, Sources, Confidence notes | Retrieval gives agents fresher context and reduces hallucination risk when answers must be grounded. | OpenAI notes apps/connectors can connect ChatGPT to third-party data sources such as email and document repositories. | Source ↗ |
| CNTechnical & BuilderCodebase Navigation | Technical & Builder | Very high | High | Medium | Onboarding, bug triage, feature planning, refactors, and code review. | Repository, Goal, Search scope | File map, Implementation notes, Risk areas | Navigation is the first technical skill; agents must know where to work before they change code. | Claude Code is described as understanding codebases, editing files, running commands, and helping developers ship faster. | Source ↗ |
| FUCore SkillsFile Understanding | Core Skills | Very high | High | Medium | Contracts, PDFs, spreadsheets, slide decks, reports, invoices, and internal documents. | Files, Task brief, Output format | Summary, Extraction, Edited artifact | Most valuable work is trapped in files; file skill makes agents useful inside existing workflows. | Anthropic provides pre-built Agent Skills for common document tasks including PowerPoint, Excel, Word, and PDF. | Source ↗ |
| DACore SkillsData Analysis & Spreadsheet Work | Core Skills | High | High | Medium | Finance models, KPI reporting, operations dashboards, CSV cleanup, and revenue analysis. | Dataset, Question, Business rules | Tables, Charts, Insights | Data analysis skill gives agents concrete business leverage because outputs can be checked and reused. | OpenAI notes ChatGPT agent can edit spreadsheets and use a terminal tool for data analysis and spreadsheet generation. | Source ↗ |
| HACore SkillsHuman Approval & Handoff | Core Skills | High | High | Low | Purchases, sending emails, production deploys, account changes, and customer escalations. | Approval policy, Escalation path, Audit needs | Approval request, Decision log, Handoff summary | Approval skill makes automation trustworthy enough for real organizations. | OpenAI emphasizes user control for ChatGPT agent when taking actions and interacting with logged-in websites. | Source ↗ |
| MCTechnical & BuilderMCP Connector Usage | Technical & Builder | High | Medium | Medium | Enterprise tool access, developer environments, databases, SaaS apps, and local workflows. | MCP server, Tool permissions, Discovery rules | Tool calls, Connected context, Action logs | MCP is becoming a practical connective layer for agent ecosystems. | Anthropic discusses code execution with MCP as a way for agents to interact with many tools more efficiently while using fewer tokens. | Source ↗ |
| SGCore SkillsSafety Guardrails | Core Skills | Very high | Medium-high | Low | Enterprise deployments, regulated workflows, customer-facing agents, and autonomous operations. | Policies, Permissions, Risk taxonomy | Blocked actions, Warnings, Audit trail | Guardrails are not decoration; they decide whether agent automation can safely move beyond demos. | OpenAI says agentic releases introduce new risks because agents can work directly with user data and websites. | Source ↗ |
| LMCore SkillsLong-Context Memory | Core Skills | High | Medium | High | Personal assistants, ongoing projects, customer history, research notebooks, and account management. | Memory policy, User preferences, Retention rules | Persistent preferences, Project context, Summaries | Memory lets an agent stop starting from zero, but it needs careful privacy and update controls. | Google highlights stateful agents and memory management in real-world agent examples, including Letta and memory hierarchy patterns. | Source ↗ |
| MCCore SkillsMulti-Agent Coordination | Core Skills | Medium-high | Medium | Medium | Research teams, software delivery, sales operations, data projects, and business process automation. | Roles, Shared state, Routing rules | Delegated tasks, Merged results, Supervisor summary | Coordination lets teams combine narrow expert agents without forcing one prompt to do everything. | OpenAI describes agents as applications that can collaborate across specialists and maintain enough state for multi-step work. | Source ↗ |
| DGWork & ProductivityDocument Generation | Work & Productivity | Very high | High | Medium | Proposals, SOPs, summaries, reports, contracts drafts, and internal documentation. | Brief, Template, Source material | Draft document, Structured sections, Citations | Document generation is one of the easiest agent wins because the output maps directly to daily work. | Anthropic lists document-focused pre-built Agent Skills for Word, PDF, PowerPoint, and Excel workflows. | Source ↗ |
| PCWork & ProductivityPresentation Creation | Work & Productivity | High | Medium-high | Medium | Sales decks, investor updates, internal reviews, training, and executive summaries. | Audience, Message, Source data | Slides, Speaker notes, Visual structure | Slide creation compresses research, writing, and formatting into one practical agent workflow. | OpenAI says ChatGPT agent can use tools to complete tasks like research, bookings, and slideshows. | Source ↗ |
| ECWork & ProductivityEmail & Calendar Operations | Work & Productivity | High | Medium-high | High | Executive assistants, sales reps, recruiting, account management, and team operations. | Mailbox/calendar access, Tone rules, Approval policy | Draft replies, Meeting options, Follow-up tasks | Email and calendar work is repetitive, high-volume, and benefits strongly from human-in-the-loop controls. | OpenAI gives examples such as summarizing an inbox or finding meeting times through connected apps. | Source ↗ |
| CSWork & ProductivityCRM & Sales Workflow | Work & Productivity | High | Medium | Medium | SDR workflows, pipeline hygiene, account plans, qualification, and renewal operations. | CRM access, Ideal customer profile, Messaging rules | Lead notes, CRM updates, Outreach drafts | Sales workflows are full of small context switches that agents can bundle into one guided action. | Google discusses agents connecting with tools such as GitHub, Slack, Google Workspace, Notion, and other services through integration layers. | Source ↗ |
| CSWork & ProductivityCustomer Support Resolution | Work & Productivity | Very high | High | Medium | Help desks, SaaS support, refunds, troubleshooting, and multilingual service. | Knowledge base, Ticket context, Escalation policy | Answer draft, Resolved ticket, Escalation summary | Support has clear ROI because repetitive questions can be answered quickly while complex cases go to humans. | Agent platforms commonly combine retrieval, tool use, and handoff patterns for production support workflows. | Source ↗ |
| MNWork & ProductivityMeeting Notes & Follow-up | Work & Productivity | High | High | Low | Sales calls, internal standups, interviews, project reviews, and customer success. | Transcript, Participants, Project context | Summary, Tasks, Follow-up message | Meeting follow-up is a clean repeatable workflow with clear acceptance criteria. | Workspace agents are positioned for repeatable workflows that otherwise require re-explaining steps and copying information between tools. | Source ↗ |
| WRWork & ProductivityWeb Research & Source Synthesis | Work & Productivity | Very high | High | Medium | Vendor research, market maps, competitive analysis, procurement, and content planning. | Research question, Recency needs, Source rules | Shortlist, Cited summary, Comparison table | Research agents save time when they cite sources and show enough structure for humans to verify. | ChatGPT agent combines research and action, including web navigation and source-connected workflows. | Source ↗ |
| FBWork & ProductivityForms & Back-Office Automation | Work & Productivity | Medium-high | Medium | High | Procurement, onboarding, HR updates, compliance checklists, and invoice routing. | Form rules, User data, Approval thresholds | Filled forms, Submission log, Exception list | Back-office workflows are repetitive but fragile, so visual browser skill plus approvals can remove real toil. | Google highlights browser-use style agents that can visually identify fields and handle multi-step form filling. | Source ↗ |
| CRWork & ProductivityContent Repurposing | Work & Productivity | High | High | Low | Marketing teams, founders, creators, product launches, and sales enablement. | Source content, Brand voice, Channels | Post variants, Email drafts, SEO snippets | Content repurposing is fast to adopt because risk is manageable and outputs are easy to review. | General-purpose agents and workspace agents are increasingly used for repeatable writing and transformation workflows. | Source ↗ |
| KBWork & ProductivityKnowledge Base Maintenance | Work & Productivity | Medium-high | Medium | Medium | Support centers, internal wikis, product docs, onboarding, and policy libraries. | Knowledge base, Change logs, Style guide | Update drafts, Gap list, Linked references | Agents get better when the knowledge they rely on stays fresh. | Google’s agent skills work emphasizes reducing knowledge gaps by packaging current docs and instructions as source-of-truth skills. | Source ↗ |
| CETechnical & BuilderCode Editing & Refactoring | Technical & Builder | Very high | High | High | Feature work, bug fixes, migrations, cleanup, and UI changes. | Codebase, Acceptance criteria, Test command | Patch, Change summary, Residual risks | Code editing is where agent value becomes tangible, but review and tests remain essential. | Anthropic describes Claude Code as an agentic coding tool that can edit files and run commands. | Source ↗ |
| TGTechnical & BuilderTest Generation & Debugging | Technical & Builder | High | Medium-high | Medium | CI failures, regression tests, bug reports, flaky workflows, and QA automation. | Failure logs, Repo context, Expected behavior | Repro steps, Test patch, Fix candidate | Debugging skill closes the loop between suggestion and verified behavior. | Coding agent workflows commonly combine filesystem access, code execution, and focused domain instructions. | Source ↗ |
| TSTechnical & BuilderTerminal & Script Execution | Technical & Builder | High | Medium-high | High | Builds, tests, migrations, data transforms, diagnostics, and deployment checks. | Command, Working directory, Safety constraints | Command output, Artifacts, Verification notes | Execution gives agents feedback from the real environment instead of guessing. | OpenAI’s ChatGPT agent system card describes a terminal tool for code execution, data analysis, and generating slides or spreadsheets. | Source ↗ |
| AITechnical & BuilderAPI Integration | Technical & Builder | High | High | Medium | Internal tools, SaaS automations, data sync, billing, notifications, and agent actions. | API docs, Credentials, Schema | Requests, Responses, Integration plan | APIs are the cleanest way for agents to act reliably in production systems. | Tool use and function calling docs emphasize connecting agents to external tools and APIs through structured interfaces. | Source ↗ |
| WOTechnical & BuilderWorkflow Orchestration | Technical & Builder | High | Medium-high | Medium | Business processes, ETL jobs, customer journeys, and cross-team operations. | Workflow graph, Triggers, Failure policy | Run state, Step outputs, Exception log | Orchestration turns isolated skills into repeatable systems. | Google notes open-source frameworks such as LangGraph, LlamaIndex, Pydantic AI, and n8n for building next-generation agents. | Source ↗ |
| DPTechnical & BuilderData Pipeline Operations | Technical & Builder | Medium-high | Medium | Medium | Analytics engineering, reporting, enrichment, ETL checks, and data quality workflows. | Data source, Schema, Validation rules | Clean data, Pipeline notes, Quality checks | Data pipeline skill gives agents a useful role in the operational layer behind dashboards and products. | Agent frameworks increasingly combine tools, knowledge, parsing, extraction, and indexing for reliable knowledge and data agents. | Source ↗ |
| OETechnical & BuilderObservability & Evaluation | Technical & Builder | Medium-high | Medium | Low | Production agents, regulated teams, support automation, sales agents, and internal copilots. | Eval set, Metrics, Logs | Scorecards, Failure clusters, Improvement plan | Without evaluation, agent improvement is vibes; with it, teams can iterate with confidence. | Anthropic’s skill best practices emphasize concise, well-structured skills tested with real usage. | Source ↗ |
| DCTechnical & BuilderDeployment & Cloud Operations | Technical & Builder | Medium | Medium | High | Agent apps, internal automation, cloud runbooks, API services, and production rollout. | Cloud target, Secrets policy, Runbook | Deployment steps, Config changes, Health checks | Deployment skill is how agent projects stop being demos and start becoming maintained systems. | Google’s Agents CLI is designed to give coding agents a machine-readable path across Google Cloud agent stack and deployment workflows. | Source ↗ |
| SClaude Code SkillsSuperpowers | Claude Code Skills | High | High | Medium | TDD-heavy engineering workflows, planning, debugging, code review, and consistent Claude Code usage across teams. | Repository, Acceptance criteria, Test strategy | Development plan, Tests-first workflow, Reviewed implementation | It turns agentic coding from one-off prompting into a repeatable engineering process. | The Superpowers GitHub repository describes it as a complete software development methodology for coding agents built on composable skills and initial instructions. | Source ↗ |
| UUClaude Code SkillsUI UX Pro Max | Claude Code Skills | Medium-high | Medium-high | Low | Landing pages, dashboards, design systems, frontend prototypes, and visual polish before handoff. | Product brief, Brand direction, Target users | Design direction, Component guidance, Frontend UI recommendations | Design taste is hard to recover after the first draft; this skill gives coding agents a stronger visual starting point. | The UI UX Pro Max GitHub repository describes an AI skill that provides design intelligence for building professional UI/UX across multiple platforms. | Source ↗ |
| CSClaude Code SkillsClaude SEO | Claude Code Skills | Medium-high | Medium | Medium | Technical SEO audits, content checks, page optimization, and search-driven website improvements. | Target URL, Keyword set, SEO goals | SEO audit, Fix list, Optimization recommendations | SEO is a perfect coding-agent workflow because it combines crawling, structured checks, content judgment, and implementation. | The Claude SEO GitHub repository describes an open-source SEO analysis plugin for Claude Code with sub-skills and specialist agents. | Source ↗ |
| ABClaude Code SkillsAgent Browser | Claude Code Skills | High | Medium-high | High | Frontend QA, form testing, web data checks, screenshot verification, and local app validation. | URL, Task goal, Safety rules | Browser actions, Screenshots, Verification notes | Browser access lets coding agents verify the actual product instead of guessing from files alone. | The Agent Browser GitHub repository describes a browser automation CLI for AI agents, including Claude Code skill installation support. | Source ↗ |
| SCClaude Code SkillsSkill Creator | Claude Code Skills | High | High | Low | Turning team SOPs, checklists, and repeated workflows into reusable Claude Code skills. | Workflow goal, Examples, Quality criteria | SKILL.md, Skill folder structure, Testing checklist | A good meta-skill compounds value because every repeated workflow can become an installable capability. | Anthropic’s official skills repository includes the Skill Creator skill for creating and packaging Claude skills. | Source ↗ |
| EDClaude Code SkillsExcalidraw Diagram | Claude Code Skills | Medium | Medium | Low | System diagrams, product flows, architecture docs, onboarding visuals, and stakeholder explanations. | Diagram goal, Entities, Style constraints | Diagram file, Visual draft, Verification notes | Diagrams help users understand agent work faster than long textual explanations. | The Excalidraw Diagram Skill GitHub repository describes a skill that gives Claude Code and other coding agents the ability to generate practical Excalidraw diagrams. | Source ↗ |
| RClaude Code SkillsRemotion | Claude Code Skills | Medium | Medium | Medium | Product videos, demos, explainers, launch assets, and motion-rich technical content. | Storyboard, Assets, Visual style | Video composition, Scene structure, Render guidance | Video generation is becoming a useful coding-agent frontier because layouts, timing, and assets can be versioned. | Remotion’s official Agent Skills documentation says its skills define best practices for AI agents such as Claude Code, Codex, and Cursor when working in Remotion projects. | Source ↗ |
| ABClaude Code SkillsArtifacts Builder | Claude Code Skills | Medium-high | Medium | Low | React prototypes, Tailwind mockups, shadcn/ui experiments, and interactive stakeholder previews. | Prototype brief, UI requirements, Framework preference | Interactive artifact, Component draft, Prototype notes | Artifacts make agent output tangible quickly, which is ideal for design and product iteration. | The Artifacts Builder skill source is available in the claude-code-templates GitHub repository as a Claude Code skill for building rich HTML artifacts. | Source ↗ |
| MCClaude Code SkillsMCP Client | Claude Code Skills | High | Medium | Medium | Tool integration, data access, SaaS actions, internal systems, and multi-tool coding workflows. | MCP server details, Permissions, Tool goal | MCP connection, Tool calls, Action results | MCP is quickly becoming the practical tool layer for coding agents and workflow agents. | Claude Code’s official MCP documentation explains how MCP connects Claude Code to external tools and exposes Claude Code tools to MCP clients. | Source ↗ |
| NClaude Code SkillsNotebookLM | Claude Code Skills | Medium | Medium | Medium | Research briefs, source comparison, citation tracking, documentation synthesis, and knowledge workflows. | Source set, Research question, Citation expectations | Grounded answer, Citations, Synthesis notes | Research skills are more useful when they preserve source grounding instead of producing unsupported summaries. | The NotebookLM Skill GitHub repository describes a Claude Code skill for querying NotebookLM and retrieving source-grounded answers. | Source ↗ |
BUILD YOUR AI TEAM
