Huginn is an open-source platform to create and manage automated agents that monitor events and perform tasks.
Best Development & Technical AI Agents (3,252)
5 subcategories · 3,252 agents · Updated October 31, 2026
Development & Technical covers the agents and infrastructure that developers use to build, run and extend AI agents. AI Platforms & Frameworks is the foundation: hosted agent environments, orchestration frameworks, multi-agent systems and protocol servers that connect models to tools and data. Development Environment and Model Services add SDKs, runtimes, model-serving engines and the plumbing for deploying agents at scale. Coding Assistance agents write, refactor, test and review code, from editor extensions to agents that work through a whole issue and open a pull request. Web Tools agents act on the internet: they browse sites, fill forms, scrape data and move information between web apps.
Choosing here is mostly an engineering decision. Check supported languages and models, whether you can self-host, and how the framework handles tool calls, memory, retries and observability once agents run in production. Look at the maturity of the project, including documentation, release history, issue response and license, since many entries are open source and some licenses restrict commercial use. For coding agents, confirm what access they get to your repositories and terminals and whether changes always go through review. For browsing agents, check how they handle logins and site terms, and keep them away from credentials they do not need. Costs usually come from the model calls an agent makes, so estimate token usage on realistic tasks.
Ads
Browse 5 Development & Technical subcategories
4,081 listings · open a subcategory to compare every agent in it
- 2,736 agents
AI Platforms & Frameworks
This category collects the platforms, runtimes and frameworks that AI agents actually run on.
- 519 agents
Coding Assistance
Explore AI coding assistants that generate and complete code, explain unfamiliar logic, find bugs, suggest refactors, write tests…
- 256 agents
Development Environment
Tools in this category give developers the groundwork for building and running AI agents: frameworks and SDKs…
- 136 agents
Model Services
Model services cover the infrastructure behind AI agents: agent SDKs, multi-agent frameworks, hosted runtimes and model-serving endpoints.
- Built a Development & Technical agent?Submit it to Creati.ai and choose the subcategories it belongs in. Listings are reviewed before they go live.
Top agents in Development & Technical
24 picks from 3,252 agents, each listed once. Ranked by traffic to the agent's own website, GitHub stars, subcategory coverage and user saves.
Hands-on course teaching creation of autonomous AI agents with Hugging Face Transformers, APIs, and custom tool integrations.
SWE-agent autonomously leverages language models to detect, diagnose, and fix issues in GitHub repositories.
A no-code AI agent platform that designs, deploys, and monitors intelligent agents to automate workflows and integrate business tools.
Freemium · $17+Visit ↗Mastra is a no-code AI agent builder that orchestrates multi-step workflows, integrates data sources, and deploys custom chatbots.
A modular SDK enabling autonomous LLM-based agents to execute tasks, maintain memory, and integrate external tools.
A decentralized AI inference marketplace connecting model owners with distributed GPU providers for pay-as-you-go serving.
GPTMe is a Python-based framework to build custom AI agents with memory, tool integration, and real-time APIs.
Agentless is an AI-powered framework that orchestrates automated code generation, execution, and validation without a dedicated agent layer.
Rags is a Python framework enabling retrieval-augmented chatbots by combining vector stores with LLMs for knowledge-based QA.
LangChain Studio offers a visual interface for building, testing, and deploying AI agents and natural language workflows.
Agent-Squad coordinates multiple specialized AI agents to decompose tasks, orchestrate workflows, and integrate tools for complex problem solving.
AppAgent uses LLM and vision to autonomously navigate and operate smartphone apps by interacting with GUIs.
OM-Agent is a no-code AI agent platform enabling custom autonomous agents to execute tasks and integrate APIs.
Desktop Commander uses AI to automate desktop tasks—launch apps, manage files, and streamline workflows via natural language commands.
AI web page summarization assistant that automatically extracts key insights and enables on-page Q&A.
MLE Agent leverages LLMs to automate machine learning operations, including experiment tracking, model monitoring, pipeline orchestration.
Agent-FLAN is an open-source AI agent framework enabling multi-role orchestration, planning, tool integration and execution of complex workflows.
RinSim is a Java-based discrete-event multi-agent simulation framework for evaluating dynamic vehicle routing, ride-sharing, and logistics strategies.
An LSP server that uses OpenAI's GPT models to automate code refactoring tasks like method extraction, variable renaming, and formatting.
Imbue is an AI agent designed to enhance conversation and collaboration through intelligent dialogue.
An AI tool that uses Anthropic Claude embeddings via CrewAI to find and rank similar companies based on input lists.
Fellou: Your next-gen cognitive browser and digital life companion.
Pay-as-you-go · $20+Visit ↗Linkup is an AI agent that automates business communications and task management.
Frequently asked questions
A framework is a library you build with: it gives you building blocks for prompts, tools, memory and multi-agent coordination, and you host the result yourself. A platform is a hosted service that runs agents for you, usually with a visual builder, integrations, monitoring and billing included. Many products in AI Platforms & Frameworks offer both. Frameworks give more control; platforms get something running faster.
Give autonomy gradually. Coding Assistance agents can complete well-defined tasks such as fixing a failing test or updating a dependency, but they can also make wide, confident changes that are wrong. Run them in a branch or sandbox, require pull requests and human review, keep production credentials out of reach, and make sure your tests cover the code they touch before relying on them for larger tasks.
Most of the cost comes from model usage, which grows with the number of steps, the length of the context and retries when something fails. Run a few realistic tasks, log tokens per task, then multiply by expected volume. Add hosting, vector storage and any per-seat or per-run platform fees. Model Services tools that support caching, or smaller models for simple steps, can lower costs considerably.
Each agent gets one combined score. Popularity carries the most weight, measured by monthly visits to the agent's own website or by GitHub stars, since many projects in this group are open source. The score also reflects how many Development & Technical subcategories the agent covers and how often Creati.ai users save it. Each agent is listed once, only agents whose main focus is this group qualify, and advertising does not affect the order.
More AI agent groups
Looking for single-purpose tools? AI Tool categories →