Anthropic says it solved the long-running AI agent problem with a new multi-session Claude SDK
Anthropic is tackling a significant challenge in the realm of artificial intelligence: the memory limitations of long-running agents. These agents, which operate based on foundation models, often struggle to retain information across multiple sessions due to the constraints of their context windows. As a result, they can forget critical instructions or previous conversations, leading to inconsistent performance, especially in complex tasks. In a recent blog post, Anthropic introduced its innovative solution through the Claude Agent SDK, which employs a two-fold approach designed to enhance agent memory and ensure smoother, more effective operations across extended projects.
The proposed solution consists of an initializer agent and a coding agent. The initializer agent sets up the working environment and logs all actions taken, creating a structured foundation for the task at hand. This preparation allows the coding agent to focus on making incremental progress during each session while ensuring that it leaves behind clear and organized updates for future sessions. This method not only addresses the issue of agents forgetting previous instructions but also mimics the effective practices of human software engineers who document their work and progress. Anthropic’s research indicates that this structured approach can significantly improve the reliability and performance of agents, particularly in tasks like full-stack web app development, where maintaining continuity is crucial.
As the demand for enhanced agent memory solutions grows, several companies, including LangChain and OpenAI, have developed their own frameworks to address similar challenges. Anthropic acknowledges that its strategy is just one of many potential solutions in this evolving field. The company is eager to explore further research avenues, particularly in understanding whether a single general-purpose coding agent or a multi-agent structure would be more effective in various contexts. Future experiments may expand beyond web app development, potentially applying these lessons to other complex domains such as scientific research or financial modeling. As the AI landscape continues to evolve, Anthropic’s advancements in agent memory could pave the way for more robust and capable AI systems in the future.
Agent memory remains a problem that enterprises want to fix, as agents forget some instructions or conversations the longer they run.
Anthropic
believes it has solved this issue for its
Claude Agent SDK
, developing a two-fold solution that allows an agent to work across different context windows.
“The core challenge of long-running agents is that they must work in discrete sessions, and each new session begins with no memory of what came before,” Anthropic wrote in
a blog post
. “Because context windows are limited, and because most complex projects cannot be completed within a single window, agents need a way to bridge the gap between coding sessions.”
Anthropic engineers proposed a two-fold approach for its Agent SDK: An initializer agent to set up the environment, and a coding agent to make incremental progress in each session and leave artifacts for the next.
The agent memory problem
Since agents are built on foundation models, they remain constrained by the limited, although continually growing, context windows. For long-running agents, this could create a larger problem, leading the agent to forget instructions and behave abnormally while performing a task.
Enhancing agent memory
becomes essential for consistent, business-safe performance.
Several methods emerged over the past year, all attempting to bridge the gap between context windows and agent memory.
LangChain
’s LangMem SDK,
Memobase
and
OpenAI
’s Swarm are examples of companies offering memory solutions. Research on agentic memory has also exploded recently, with proposed
frameworks like Memp
and the
Nested Learning Paradigm
from
Google
offering new alternatives to enhance memory.
Many of the current memory frameworks are open source and can ideally adapt to different large language models (LLMs) powering agents. Anthropic’s approach improves its Claude Agent SDK.
How it works
Anthropic identified that even though the Claude Agent SDK had context management capabilities and “should be possible for an agent to continue to do useful work for an arbitrarily long time,” it was not sufficient. The company said in its blog post that a model
like Opus 4.5
running the Claude Agent SDK can “fall short of building a production-quality web app if it’s only given a high-level prompt, such as ‘build a clone of claude.ai.’”
The failures manifested in two patterns, Anthropic said. First, the agent tried to do too much, causing the model to run out of context in the middle. The agent then has to guess what happened and cannot pass clear instructions to the next agent. The second failure occurs later on, after some features have already been built. The agent sees progress has been made and just declares the job done.
Anthropic researchers broke down the solution: Setting up an initial environment to lay the foundation for features and prompting each agent to make incremental progress towards a goal, while still leaving a clean slate at the end.
This is where the two-part solution of Anthropic’s agent comes in. The initializer agent sets up the environment, logging what agents have done and which files have been added. The coding agent will then ask models to make incremental progress and leave structured updates.
“Inspiration for these practices came from knowing what effective software engineers do every day,” Anthropic said.
The researchers said they added testing tools to the coding agent, improving its ability to identify and fix bugs that weren’t obvious from the code alone.
Future research
Anthropic noted that its approach is “one possible set of solutions in a long-running agent harness.” However, this is just the beginning stage of what could become a wider research area for many in the AI space.
The company said its experiments to boost long-term memory for agents haven’t shown whether a single general-purpose coding agent works best across contexts or a multi-agent structure.
Its demo also focused on full-stack web app development, so other experiments should focus on generalizing the results across different tasks.
“It’s likely that some or all of these lessons can be applied to the types of long-running agentic tasks required in, for example, scientific research or financial modeling,” Anthropic said.
