Search “Agents”

NVIDIA expands open-source Agent Toolkit and launches Agent Knowledge Foundation

On October 21, 2026, NVIDIA announced the expansion of its open-source Agent Toolkit and the launch of the Agent Knowledge Foundation. According to information published by the NVIDIA Newsroom that day, the expansion also adds the OpenShell secure runtime, the NemoClaw blueprint, and tools such as NeMo Relay and NeMo Compass. The Agent Knowledge Foundation includes CUDA-X libraries and the new Auto-Ontology library, and integrates the Nemotron Retriever and the Kumo family of open models, allowing enterprises to use it without migrating data or rebuilding existing workflows. The same source said that companies such as Palantir, Databricks, and Snowflake have already integrated the toolkit into their respective platforms. There are currently no subsequent independent developments.

Shopify shares how Sidekick uses a post-trained open-source model as the GraphQL sub-agent

On October 21, 2026, the Shopify engineering team, in a talk at NVIDIA GTC Berlin 2026, shared the model division-of-labor plan for its e-commerce assistant Sidekick: a frontier model handles conversation planning, while a self-deployed open-source model handles the GraphQL sub-agent. The team chose the MoE model Nemotron 3.5 Lightning, with 30 billion total parameters and 3 billion parameters activated per token; after 6 rounds of post-training and 61,000 repair samples, the review score rose from 37.7 out of the box to 74.5, slightly above the frontier model's 73. On GraphQL tasks, the median response was 0.87 seconds, versus 6.50 seconds for the frontier model API, at roughly 90% lower cost. Currently this talk is a single-source report, with no follow-up developments yet.

English translation is being prepared.

On October 21, 2026, the Shopify engineering team, in a talk at NVIDIA GTC Berlin 2026, shared the model division-of-labor plan for its e-commerce assistant Sidekick: a frontier model handles conversation planning, while a self-deployed open-source model handles the GraphQL sub-agent. The team chose the MoE model Nemotron 3.5 Lightning, with 30 billion total parameters and 3 billion parameters activated per token; after 6 rounds of post-training and 61,000 repair samples, the review score rose from 37.7 out of the box to 74.5, slightly above the frontier model's 73. On GraphQL tasks, the median response was 0.87 seconds, versus 6.50 seconds for the frontier model API, at roughly 90% lower cost. Currently this talk is a single-source report, with no follow-up developments yet.

View story:Shopify shares how Sidekick uses a post-trained open-source model as the GraphQL sub-agent

Li Biao proposed that there is no Personal Agent, only Context plus Agent

On October 21, 2026, Li Biao put forward the view: there is no independent Personal Agent, only Context plus Agent. The related article further explained that the same Agent connected to work context is a Work Agent, and connected to life context is a Personal Agent. The article also mentioned that domestic Work Agents already have tens of millions of users, while life context is scattered across super apps; Alipay has opened over 300 services, and Meituan has only opened errand running. Domestic vendors have therefore turned to Agent phones, while the United States mostly takes the cloud route; the long-term barrier for Agents lies in the Context flywheel, but authorization boundaries, error-correction records, and the like are currently difficult to migrate with users.

English translation is being prepared.

On October 21, 2026, Li Biao put forward the view: there is no independent Personal Agent, only Context plus Agent. The related article further explained that the same Agent connected to work context is a Work Agent, and connected to life context is a Personal Agent. The article also mentioned that domestic Work Agents already have tens of millions of users, while life context is scattered across super apps; Alipay has opened over 300 services, and Meituan has only opened errand running. Domestic vendors have therefore turned to Agent phones, while the United States mostly takes the cloud route; the long-term barrier for Agents lies in the Context flywheel, but authorization boundaries, error-correction records, and the like are currently difficult to migrate with users.

View story:Li Biao proposed that there is no Personal Agent, only Context plus Agent

Nvidia adds Agent Knowledge Foundation to Agent Toolkit

At GTC Berlin, Nvidia added the Agent Knowledge Foundation to its Agent Toolkit. The component packages the Nemotron Retriever models, a new Auto-Ontology library, the Kumo family of models for structured data, and CUDA-X data processing libraries; data remains in place and is governed by enterprises' existing permissions. Nvidia says Nemotron Retriever improved search relevance by 65% in tests in the open-source OpenSearch environment, SAP's fine-tuned reranker runs nearly three times faster, and ServiceNow's reranking offers better quality with nearly 25% lower latency. Auto-Ontology provides a REST API and an MCP server for agents to query. Nvidia says it does not compete with data platforms such as Snowflake, but partners with them. The news was reported by The New Stack; there are no subsequent developments yet.

View story:Nvidia adds Agent Knowledge Foundation to Agent Toolkit

Microsoft proposes using Guardian Agents to monitor enterprise AI agents

Microsoft has proposed a plan to use Guardian Agents to monitor and regulate the actions of enterprise AI agents. Hayete Gallot, Executive Vice President of Microsoft's security division, said in a blog post that Guardian Agents should be deployed to monitor and intervene before AI agents execute actions, and stated that frontier models cannot be trusted no matter where they come from. The plan intends to bring enterprise agents under centralized management, assigning each agent an identity, container, and permission policy, then combining identity, data, cloud, endpoint, network, and runtime signals to assess risk in real time, with red team and blue team agents intervening when risk rises. The plan will be implemented through Microsoft's Project Perception security system announced in July. The plan is still at the proposal stage, and there are no public facts about further implementation progress.

View story:Microsoft proposes using Guardian Agents to monitor enterprise AI agents

English translation is being prepared.

English translation is being prepared.

View story:English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

View story:English translation is being prepared.

OpenRouter says agentic usage has grown 34-fold since February

OpenRouter stated that agentic usage has grown 34-fold since early February 2026. This claim was published by OpenRouter on October 21, 2026, and is first-party information from the company. The same source also said that, on a weekly basis, agentic usage is currently about 8 times human usage. There are no subsequent developments at present, and the status remains at the above disclosure.

View story:OpenRouter says agentic usage has grown 34-fold since February

English translation is being prepared.

Massimo Re Ferre argues that trust determines whether Agent coding gets adopted

On October 20, 2026, Massimo Re Ferre, AWS Senior Principal Product Manager, proposed that the degree to which a team trusts AI agents determines how much coding autonomy they are granted. In his view, the key for teams moving from individual coding assistants to agent development platforms lies not in model capability, but in whether they can verify agent output at low cost; trust rises with verification capability, not with benchmarks. He also noted that agents have evolved from line-level completion to autonomously selecting tasks, and that costs have shifted from fixed per-seat pricing to billing based on task activity—about 80% of developers spend only 20% of the budget, while a small number of heavy users consume the rest—so the focus should be on the cost per unit of effective work rather than seat price. These views come from his talk at DevCon and an article on the Tessl blog; there are currently no further developments.

Nvidia adds Agent Knowledge Foundation to Agent Toolkit

At GTC Berlin, Nvidia added the Agent Knowledge Foundation to its Agent Toolkit. The component packages the Nemotron Retriever models, a new Auto-Ontology library, the Kumo family of models for structured data, and CUDA-X data processing libraries; data remains in place and is governed by enterprises' existing permissions. Nvidia says Nemotron Retriever improved search relevance by 65% in tests in the open-source OpenSearch environment, SAP's fine-tuned reranker runs nearly three times faster, and ServiceNow's reranking offers better quality with nearly 25% lower latency. Auto-Ontology provides a REST API and an MCP server for agents to query. Nvidia says it does not compete with data platforms such as Snowflake, but partners with them. The news was reported by The New Stack; there are no subsequent developments yet.

OpenRouter says agentic usage has grown 34-fold since February

OpenRouter stated that agentic usage has grown 34-fold since early February 2026. This claim was published by OpenRouter on October 21, 2026, and is first-party information from the company. The same source also said that, on a weekly basis, agentic usage is currently about 8 times human usage. There are no subsequent developments at present, and the status remains at the above disclosure.

Rohan Paul says Meta's AI agent Muse reached 10.2 million downloads in the US in 40 days

Rohan Paul offered the assessment that Meta's AI agent Muse reached 10.2 million downloads in the US within 40 days of launch. The data he cited also shows that historically only 11 apps have reached this pace, most of them mobile games. On retention, Muse's US 7-day retention rate is 30% and its 14-day retention rate is 23%, compared with competitor averages of 19% and 16%, respectively. On October 16, its daily active users reached 2.5 million, up 15% in a week, but still less than a quarter of its US downloads. At present, this assessment is seen only from Rohan Paul as a single source, with no other independent sources or official evidence to cross-verify it.

Rohan Paul says Meta's AI agent Muse reached 10.2 million US downloads in 40 days

Rohan Paul offered the assessment that Meta's AI agent Muse reached 10.2 million downloads in the US within 40 days of launch. The data he cited also shows that historically only 11 apps have reached this pace, most of them mobile games. On retention, Muse's US 7-day retention rate is 30% and its 14-day retention rate is 23%, compared with competitor averages of 19% and 16%, respectively. On October 16, its daily active users reached 2.5 million, up 15% in a week, but still less than a quarter of its US downloads. At present, this assessment is seen only from Rohan Paul as a single source, with no other independent sources or official evidence to cross-verify it.

View story:Rohan Paul says Meta's AI agent Muse reached 10.2 million downloads in the US in 40 days

Massimo Re Ferre: Trust determines whether agentic coding gets adopted

On October 20, 2026, Massimo Re Ferre, AWS Senior Principal Product Manager, proposed that the degree to which a team trusts AI agents determines how much coding autonomy they are granted. In his view, the key for teams moving from individual coding assistants to agent development platforms lies not in model capability, but in whether they can verify agent output at low cost; trust rises with verification capability, not with benchmarks. He also noted that agents have evolved from line-level completion to autonomously selecting tasks, and that costs have shifted from fixed per-seat pricing to billing based on task activity—about 80% of developers spend only 20% of the budget, while a small number of heavy users consume the rest—so the focus should be on the cost per unit of effective work rather than seat price. These views come from his talk at DevCon and an article on the Tessl blog; there are currently no further developments.

View story:Massimo Re Ferre argues that trust determines whether Agent coding gets adopted

Microsoft proposes using Guardian Agents to monitor enterprise AI agents

Microsoft has proposed a plan to use Guardian Agents to monitor and regulate the actions of enterprise AI agents. Hayete Gallot, Executive Vice President of Microsoft's security division, said in a blog post that Guardian Agents should be deployed to monitor and intervene before AI agents execute actions, and stated that frontier models cannot be trusted no matter where they come from. The plan intends to bring enterprise agents under centralized management, assigning each agent an identity, container, and permission policy, then combining identity, data, cloud, endpoint, network, and runtime signals to assess risk in real time, with red team and blue team agents intervening when risk rises. The plan will be implemented through Microsoft's Project Perception security system announced in July. The plan is still at the proposal stage, and there are no public facts about further implementation progress.