Agentic frameworks
Also written: AI Agent Framework, Agent Framework, AgentCore, Agent factory, Agentic foundation
Frameworks that help build AI systems where agents perceive, decide, and act autonomously to achieve goals.
Agentic frameworks are tools and structures that help developers create AI systems capable of acting independently. Think of them like a set of instructions and building blocks for an AI that can figure out what to do and then do it.
For example, an AI agent built with such a framework might be tasked with planning a trip. It would need to understand the goal (plan a trip), gather information (flight prices, hotel availability), make decisions (book this flight, that hotel), and then take action (make the booking). The framework provides the underlying structure for this perception, decision-making, and action loop.
Agentic frameworks provide the architecture and tools for developing autonomous agents. These agents are designed to perceive their environment through sensors or data inputs, reason about the best course of action based on their goals and learned policies, and execute actions in the physical or digital world. Key components often include planning modules, memory systems, tool-use capabilities, and execution engines. Tradeoffs involve complexity in design and debugging versus the potential for sophisticated, emergent behaviors and task automation.
Recent advancements in large language models have made it more feasible to build sophisticated AI agents. Companies are now exploring how these agents can automate complex workflows, as seen with Google Gemini's AI agents for academic tasks and Anthropic's Bloom for AI evaluations.
- Introducing Microsoft 365 G7: AI tools for government teamsMicrosoft · Microsoft Source ·
- OpenAI Cookbook: Add ChatGPT Agents SDK and Amazon Bedrock AgentCore cookbook (#3057)OpenAI · OpenAI Cookbook ·
- This company has more AI agents than employeesMicrosoft · Microsoft on YouTube ·
- OpenAI Cookbook: Remove unreviewed AgentCore companion link (#2945)OpenAI · OpenAI Cookbook ·
- Improving the academic workflow: Introducing two AI agents for better figures and peer reviewGoogle · Google Research Blog ·
- OpenClaw: The Viral AI Agent that Broke the Internet - Peter Steinberger | Lex Fridman Podcast #491Anthropic · Lex Fridman ·
- Introducing Bloom: an open source tool for automated behavioral evaluationsAnthropic · Anthropic Research ·
- Using OpenAI o1 for financial analysisOpenAI · OpenAI News ·
- Anthropic Claude
- Anthropic is developing tools like Bloom to evaluate the behavior of AI systems, which can be seen as a form of agentic evaluation.
- Google Gemini
- Google Gemini is launching AI agents designed to assist with specific tasks in the academic workflow, demonstrating practical applications of agentic frameworks.
- Orchestration frameworks
- These focus on managing and coordinating multiple AI m
- Task automation software
- This is a broader category of software that automates
- Prompt engineering
- This involves crafting inputs for AI models, while agentic f
The concept of agentic systems has roots in artificial intelligence research dating back decades, with modern agentic frameworks often built upon advancements in large language models and reinforcement learning.
Try using a chatbot like ChatGPT or Claude and ask it to perform a multi-step task, such as planning a simple itinerary for a weekend trip. Observe how it breaks down the request and asks clarifying questions, mimicking an agent's process.
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- PromptRole-play a student struggling with math for AI tutor simulationSimulate a student's math struggles and uncertainty for AI tutor evaluation. · 5 min
- TemplateReusable workflow templates for healthcare teamsTeams can create and share templates for common workflows like documentation or prior authorizations. · 15 min
- TemplateTemplate for building an AI tool tailored to your workA structured approach to defining and building a custom AI tool for specific business needs. · 15 min
- TemplateEvals CSV prompt list for agent skillsA CSV file template for listing prompts used in evaluating agent skills. · 20 min
- HeuristicDouble coding agent task completion with docs linksProviding coding agents with a documentation link significantly increases their task completion rate. · 2 min
Questions people ask
- What is the main goal of an agentic framework?
- The main goal is to enable AI systems to perceive their environment, make decisions, and take actions autonomously to achieve specific objectives.
- How do agentic frameworks differ from regular AI programs?
- Unlike regular AI programs that follow predefined scripts, agentic systems can adapt, learn, and make independent decisions to achieve goals in dynamic environments.
First seen on Subvolts Feb 13, 2025. Updated Oct 12, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.