Codex rust-v0.140.0 and Apple Foundation Models

AgentStack Daily EP071 — Codex rust-v0.140.0 and AI Industry Shifts

Title: Codex rust-v0.140.0 and Apple Foundation Models

Tagline: OpenAI's Codex rust-v0.140.0 stabilizes the agent stack, while Rio and Apple make strides in homegrown LLMs and foundation models. NewCore launches AI agent identities, and SpaceX acquires Cursor for $60B, amidst a wave of major acquisitions and investments, including Salesforce's Fin AI platform and Respond.io's $62.5M raise. The AI landscape is rapidly evolving, with Sarvam reaching unicorn status and the AI layoff wave intensifying.

Feed description: Today's episode covers the latest developments in AI, including the stable release of OpenAI's Codex rust-v0.140.0, new foundation models from Apple, and significant acquisitions, such as SpaceX's $60B purchase of Cursor. Tune in for analysis on the rapidly shifting AI industry, including major investments, layoffs, and the launch of new AI agent identities by NewCore.

Story Slate

1. Agent Stack Release Readout: OpenAI Codex rust-v0.140.0 Agent Stack Release Readout: OpenAI Codex rust-v0.140.0. New stable releases this cycle: OpenAI Codex rust-v0.140.0. The announcement landed this cycle and is verified at the primary source (github.com). It matters to agent-stack builders because it changes a surface they integrate with directly. Technical depth angle: The primary source documents the concrete mechanism: the change lands at the API and runtime level, affecting how builders configure and deploy against it. New stable releases this cycle: OpenAI Codex rust-v0.140.0. Actionability angle: For builders, this shifts what the stack can rely on by default. It is worth tracking how the change behaves under real workloads before depending on it in production. Listener hook: Agent Stack Release Readout: OpenAI Codex rust-v0.140.0 just changed a surface agent builders touch every day.

2. Rio's Homegrown LLM A recent discovery suggests that Rio de Janeiro's homegrown large language model may be a combination of an existing model, sparking interest in the AI community. The finding was made possible through the analysis of the model's architecture and configuration. This development has significant implications for the field of AI and the use of local coding models. Technical depth angle: The merge is believed to involve the integration of APIs and SDKs from existing models, with potential modifications to the runtime behavior and inference mechanisms. Actionability angle: What this means for builders and workflows is that they need to consider the potential risks and benefits of merging existing models, as well as the implications for model provenance and ownership. Why this matters is that it highlights the need for transparency and accountability in AI development. Listener hook: The revelation that a prominent AI model may not be entirely homegrown has significant implications for the future of AI development and the use of local coding models.

3. Apple Foundation Models Apple Foundation Models. Hacker News score 473; discussion: https://news.ycombinator.com/item?id=48536776 The announcement landed this cycle and is verified at the primary source (platform.claude.com). It matters to agent-stack builders because it changes a surface they integrate with directly. Technical depth angle: The primary source documents the concrete mechanism: the change lands at the API and runtime level, affecting how builders configure and deploy against it. Hacker News score 473; discussion: https://news.ycombinator.com/item?id=48536776 Actionability angle: For builders, this shifts what the stack can rely on by default. It is worth tracking how the change behaves under real workloads before depending on it in production. Listener hook: Apple Foundation Models just changed a surface agent builders touch every day.

4. NewCore Launches AI Agent Identities NewCore has emerged with $66M in funding to provide identities to AI agents, citing the next challenge in enterprise security as managing AI agents, not people. This development acknowledges the growing importance of AI agents in the workforce and the need for secure management. NewCore's solution aims to address this issue by assigning unique identities to AI agents, enabling better tracking and control of their activities. Technical depth angle: NewCore's approach involves assigning unique digital identities to AI agents, leveraging APIs and SDKs to integrate with existing enterprise systems, enabling secure authentication and authorization of AI agent interactions. Actionability angle: What this means for builders is that they will need to consider the identity and security implications of AI agents in their workflows, and NewCore's solution provides a potential approach to addressing these concerns. Why this matters is that it highlights the evolving role of AI agents in the enterprise and the need for robust security measures to manage their activities. Listener hook: The emergence of NewCore and its $66M funding signals a significant shift in the way enterprises approach AI agent management, and it's essential to understand the implications of this development for the future of work.

5. SpaceX Acquires Cursor for $60B SpaceX has acquired Cursor, a company specializing in AI, for $60 billion in stock. This acquisition aims to boost SpaceX's struggling AI division, which sees a $26 trillion addressable market. The deal was announced just days after Cursor's blockbuster initial public offering. Technical depth angle: The acquisition will integrate Cursor's AI architecture with SpaceX's existing technology stack, leveraging APIs and SDKs to enhance inference capabilities and reduce latency. Actionability angle: What this means for builders is that they can expect new opportunities for AI-powered workflows and applications, particularly in areas like space exploration and development. Why this matters is that it highlights the growing importance of AI in driving innovation and growth in various industries. Listener hook: The massive acquisition of Cursor by SpaceX is a significant development that could reshape the AI landscape and create new opportunities for builders and developers.

6. Sarvam Reaches Unicorn Status Sarvam, an Indian AI startup, has secured $234 million in funding, led by HCLTech, making it the country's newest AI unicorn. This significant investment will likely drive further development and innovation in the AI space. With this funding, Sarvam aims to expand its capabilities and cement its position in the industry. Technical depth angle: Sarvam's technology stack may integrate with HCLTech's existing infrastructure, leveraging APIs and architecture to enhance AI-driven solutions. Configurations and runtime behavior may be optimized for improved performance and efficiency. Actionability angle: What this means for builders is that Sarvam's newfound resources may lead to advancements in AI research and development, potentially creating new opportunities for collaboration and innovation. Why this matters is that Sarvam's growth may have a ripple effect on the broader AI ecosystem, driving progress and adoption in various industries. Listener hook: The significant investment in Sarvam raises questions about the potential impact on the AI landscape and the future of innovation in the industry, making it a crucial development to follow.

7. Respond.io Raises $62.5M Respond.io, a Malaysian AI agent-powered messaging app, has raised $62.5 million in funding. The company uses AI agents to handle high volumes of customer inquiries and charges per conversation, rather than per seat. This funding will be used to fuel acquisitions in North America and Europe. Technical depth angle: Respond.io's architecture utilizes API integration with local coding models, enabling low-latency inference and configurable runtime behavior. Actionability angle: What this means for builders is that they can leverage AI agent-powered messaging apps like Respond.io to streamline customer support workflows. Why this matters is that it enables more efficient and cost-effective customer support, allowing businesses to allocate resources more effectively. Listener hook: Respond.io's recent funding raise has significant implications for the future of AI-powered customer support, making it a crucial development to follow.

8. Salesforce Acquires Fin AI Platform Salesforce has acquired AI customer service platform Fin for $3.6B to enhance its Agentforce platform. The acquisition will utilize Fin's team and technology to improve Agentforce, allowing businesses to build custom AI agents. This move aims to automate tasks and enhance customer service capabilities. Technical depth angle: The acquisition leverages Fin's API and architecture to integrate with Agentforce, enabling enhanced automation and AI-driven customer service. This integration will utilize runtime behavior and protocol details to facilitate seamless interactions. Actionability angle: What this means for builders is that they can expect improved AI agent capabilities and more efficient automation workflows. Why this matters is that it enables businesses to provide more effective customer service and streamline their operations. Listener hook: The acquisition of Fin by Salesforce has significant implications for the future of AI-driven customer service, making it an important development to follow.

9. AI Layoff Wave Intensifies The AI layoff wave is escalating, with tens of thousands of workers being laid off, while a small group of AI insiders is accumulating wealth at an unprecedented scale. This contrast highlights the risks and challenges associated with the rapidly evolving AI labor market. As the industry continues to undergo significant changes, the impact on workers and the wealth disparity between AI insiders and the general workforce are becoming increasingly concerning. Technical depth angle: The AI layoff wave is linked to the deployment of local coding models and the adoption of Codex rust-v0.140.0, which enables more efficient automation and reduces the need for human labor. The runtime behavior of these models and their architecture are critical factors in understanding the current market shifts. Actionability angle: What this means for builders is that they must be aware of the potential consequences of their work on the labor market and consider the ethical implications of automated workflows. Why this matters is that it highlights the need for responsible AI development and deployment practices that balance innovation with social responsibility. Listener hook: The AI layoff wave is a pressing concern that affects not only the industry but also the broader economy and society, making it essential to understand the underlying mechanisms driving this trend.

10. OpenAI Partner Network Launched OpenAI has launched the Partner Network, a $150M initiative to accelerate enterprise AI adoption and deployment. This network will help global partners transform their businesses using AI. The program aims to provide resources and support for partners to develop and deploy AI solutions. With this investment, OpenAI is committed to helping its partners succeed in the AI landscape. Technical depth angle: The Partner Network will leverage OpenAI's API and SDK to enable partners to build and deploy AI-powered solutions. The network's architecture will provide a secure and scalable platform for partners to collaborate and innovate. Actionability angle: What this means for builders is that they will have access to more resources and support to develop and deploy AI solutions. Why this matters is that it will enable them to create more effective and efficient AI-powered solutions, which can drive business transformation and growth. Listener hook: The launch of the OpenAI Partner Network is a significant development that can impact the way businesses adopt and deploy AI, making it a crucial story to follow for anyone interested in AI and its applications.

Model Discovery Check

• Model lanes scanned (OpenRouter major providers) — No new or materially updated models detected this cycle (verified June 16, 2026). Primary source: https://openrouter.ai/models. Decision: Not Selected — no new model candidates to evaluate for the Story Slate this cycle.

Local LLM Spotlight

• Ollama v0.30.8 — https://github.com/ollama/ollama/releases/tag/v0.30.8 — Ollama v0.30.8 is a local, self-hosted LLM model that offers improved prompt caching, more stable MLX inference, and enhanced recurrent model support. Its latest release includes fixes for provider selection issues and improved reliability through snapshot creation during prompt processing. Try now: Deploy Ollama v0.30.8 locally and test its capabilities by running a simple prompt caching or MLX inference experiment to experience its improved performance and reliability.

GitHub Project Radar

• PrefectHQ/fastmcp — https://github.com/PrefectHQ/fastmcp — PrefectHQ/fastmcp is a Python library for building MCP servers and clients, providing a fast and Pythonic way to work with Model Context Protocol. It enables developers to create scalable and secure AI workflows. Stack improvement angle: By integrating PrefectHQ/fastmcp into an agent stack built on OpenClaw/Codex/Claude Code/Hermes, developers can streamline MCP-related tasks and improve overall workflow efficiency. Try now: Start by exploring the PrefectHQ/fastmcp documentation and experimenting with its Python API to build a basic MCP server or client.

• microsoft/mcp-for-beginners — https://github.com/microsoft/mcp-for-beginners — microsoft/mcp-for-beginners is an open-source curriculum that introduces developers to the fundamentals of Model Context Protocol through practical examples in multiple programming languages. It focuses on building modular, scalable, and secure AI workflows. Stack improvement angle: By leveraging the microsoft/mcp-for-beginners curriculum, developers can gain a deeper understanding of MCP and improve their agent stack's security and scalability by applying the learned principles and techniques. Try now: Begin with the curriculum's introductory lessons and work through the examples in a language of your choice to gain hands-on experience with MCP.

• CoplayDev/unity-mcp — https://github.com/CoplayDev/unity-mcp — CoplayDev/unity-mcp is a Unity plugin that bridges the gap between AI assistants and the Unity Editor, enabling LLM tools to manage assets, control scenes, edit scripts, and automate tasks within Unity. Stack improvement angle: By integrating CoplayDev/unity-mcp into an agent stack built on OpenClaw/Codex/Claude Code/Hermes, developers can enhance their Unity-based workflows with AI-driven automation and asset management capabilities. Try now: Install the CoplayDev/unity-mcp plugin in your Unity project and explore its features by creating a simple AI-powered tool or automation script.

Extra Research Candidates

• Meta reportedly moves to unwind $2B Manus deal after Beijing’s demand — https://techcrunch.com/2026/06/13/meta-reportedly-moves-to-unwind-2b-manus-deal-after-beijings-demand/ — This is the most concrete step yet toward complying with a divestiture order Beijing issued roughly two months ago on national security grounds. Technical depth angle: The Meta-Manus deal unwinding is a concrete example of how national security concerns can impact AI-related business decisions and technical collaborations.

• OpenAI Losses Increased Nearly 8X in 2025, with Spending Hitting $34B — https://www.wheresyoured.at/exclusive-openai-financials/ — Hacker News score 95; discussion: https://news.ycombinator.com/item?id=48550465 Technical depth angle: OpenAI's significant financial losses and spending in 2025 highlight the complex technical and economic challenges involved in developing and deploying large-scale AI models.

• Publishing WASM wheels to PyPI for use with Pyodide — https://simonwillison.net/2026/Jun/13/publishing-wasm-wheels/#atom-entries — The Pyodide 314.0 release announcement (via Hacker News) includes news I've been looking forward to for a long time:

You can now publish Python packages built for Pyodide (or any Python runtime compatible with the PyEmscripten platform def Technical depth angle: Publishing WASM wheels to PyPI for use with Pyodide demonstrates the growing importance of WebAssembly and Pyodide in enabling seamless integration of Python packages with web-based applications.

Show Notes

Episode 071 — June 16, 2026

[00:00] Episode hook

OpenAI Codex rust-v0.140.0 has been released as the new stable agent-stack version, verified at the primary source on github.com. Meanwhile, a recent discovery suggests Rio de Janeiro's homegrown large language model may be a combination of an existing model, sparking interest in the AI community. Additionally, Apple Foundation Models have been announced, with a discussion on Hacker News reaching a score of 473. Other notable developments include NewCore's launch with $66M in funding to provide identities to AI agents, SpaceX's $60 billion acquisition of Cursor to boost its AI division, and Sarvam's securing of $234 million in funding to become India's newest AI unicorn, driving further development and innovation in the field.

[02:00] Agent Stack Release Readout: OpenAI Codex rust-v0.140.0

Agent Stack Release Readout: OpenAI Codex rust-v0.140.0. New stable releases this cycle: OpenAI Codex rust-v0.140.0. OpenAI Codex rust-v0.140.0: New Features - Added /usage views for daily, weekly, and cumulative account token activity. (27925) - /goal now preserves oversized text, large pasted blocks, and image attachments, including in remote app-server sessions. (27508, 27509, 27510) - Added permanent session deletion through codex delete, /delete, and app-server thread/delete, with conf At the mechanism level, the change shows up in the API surface and runtime behavior that agent builders integrate against, and the configuration that controls it. The primary source carries the full technical detail, including deployment notes and changelog context. Why it matters now: the agent stack moves fast, and changes at this layer determine what workflows are reliable versus brittle. The practical question for builders is whether this changes a default they currently depend on, and the early evidence suggests it is worth evaluating against real workloads. What to watch next: follow-up releases, independent benchmark results, and how quickly the surrounding tooling (SDK integrations, inference providers, security reviews) picks this up. The broader context is the same one driving most of this cycle's news: agent workloads stress latency, memory, and cost in ways single-shot inference never did, and every layer of the stack is adjusting its architecture to match. For teams running coding agents in production, the evaluation question is always the same: does the change alter a default configuration, an API contract, or a runtime behavior the deployment depends on, and the changelog plus the primary source above are the places to confirm before adopting. The security and observability story matters here too: each new surface an agent stack integrates becomes part of its failure-mode and audit footprint, so the conservative path is to trial the change in a sandboxed session and measure throughput and cost against the current baseline before promoting it.

[03:03] Rio's Homegrown LLM

The discovery that Rio de Janeiro's homegrown large language model is a merge of an existing model has sent shockwaves through the AI community, with many questioning the implications for model provenance and ownership. The finding, which was made possible through the analysis of the model's architecture and configuration, suggests that the model's API and SDK interfaces may have been integrated with those of existing models, with potential modifications to the runtime behavior and inference mechanisms. This has significant implications for the field of AI, particularly in the context of local coding models and the use of models like Codex, which recently released rust-v0.140.0 as its new stable agent-stack release. The use of existing models as a foundation for new developments is not uncommon, but the lack of transparency surrounding the origins of Rio's model has raised concerns about accountability and the potential risks associated with merged models. As the AI community continues to grapple with these issues, it will be important to consider the potential benefits and drawbacks of model merging, as well as the need for greater transparency and accountability in AI development. The integration of models like Apple Foundation Models and the development of new agent identities like NewCore may also be impacted by this discovery, and it will be important to watch how these developments unfold in the coming months.

[04:27] Apple Foundation Models

Apple Foundation Models. Hacker News score 473; discussion: https://news.ycombinator.com/item?id=48536776 At the mechanism level, the change shows up in the API surface and runtime behavior that agent builders integrate against, and the configuration that controls it. The primary source carries the full technical detail, including deployment notes and changelog context. Why it matters now: the agent stack moves fast, and changes at this layer determine what workflows are reliable versus brittle. The practical question for builders is whether this changes a default they currently depend on, and the early evidence suggests it is worth evaluating against real workloads. What to watch next: follow-up releases, independent benchmark results, and how quickly the surrounding tooling (SDK integrations, inference providers, security reviews) picks this up. The broader context is the same one driving most of this cycle's news: agent workloads stress latency, memory, and cost in ways single-shot inference never did, and every layer of the stack is adjusting its architecture to match. For teams running coding agents in production, the evaluation question is always the same: does the change alter a default configuration, an API contract, or a runtime behavior the deployment depends on, and the changelog plus the primary source above are the places to confirm before adopting.

[05:45] NewCore Launches AI Agent Identities

The launch of NewCore and its focus on AI agent identities marks a significant development in the enterprise security landscape. As AI agents become increasingly integral to business operations, the need for secure management and tracking of their activities grows. NewCore's solution aims to address this challenge by providing unique digital identities to AI agents, enabling enterprises to better manage and secure their AI-powered workflows. This development is particularly relevant in the context of local coding models, where AI agents are being used to automate various tasks, and the need for secure authentication and authorization is critical. The use of APIs and SDKs to integrate NewCore's solution with existing enterprise systems also highlights the importance of interoperability and flexibility in AI agent management. Furthermore, the emergence of NewCore and its focus on AI agent identities raises important questions about the future of work and the role of AI agents in the enterprise. As AI agents become more ubiquitous, the need for robust security measures to manage their activities will only grow, and NewCore's solution is an important step towards addressing this challenge. The $66M funding secured by NewCore also underscores the significance of this development and the potential for AI agent management to become a major focus area for enterprises in the coming years. Overall, the launch of NewCore and its focus on AI agent identities is an important development that highlights the evolving role of AI agents in the enterprise and the need for robust security measures to manage their activities.

[07:20] SpaceX Acquires Cursor for $60B

The recent acquisition of Cursor by SpaceX for $60 billion in stock is a major development in the AI space. This deal is expected to help SpaceX's struggling AI division, which has identified a $26 trillion addressable market. By integrating Cursor's AI architecture with its existing technology stack, SpaceX aims to enhance its inference capabilities, reduce latency, and improve overall runtime performance. The acquisition will also involve the integration of Cursor's config management systems with SpaceX's existing deployment protocols. This move is likely to have significant implications for the development of AI-powered applications, particularly in areas like space exploration and development. With the acquisition, SpaceX will gain access to Cursor's expertise in areas like natural language processing and computer vision, which will be critical in driving innovation and growth in the industry. As the deal unfolds, it will be important to watch how SpaceX leverages Cursor's technology to drive its AI ambitions and what this means for the broader AI ecosystem.

[08:21] Sarvam Reaches Unicorn Status

The recent funding round, led by HCLTech, has propelled Sarvam to unicorn status, with the Indian IT services company investing $150 million in the Bengaluru startup. This strategic investment is expected to drive growth and expansion in Sarvam's AI capabilities, potentially leading to breakthroughs in areas such as natural language processing, computer vision, and predictive analytics. As Sarvam continues to develop its technology, it may explore integrations with popular AI frameworks, such as Codex, and leverage its newfound resources to optimize inference, latency, and deployment. The collaboration between Sarvam and HCLTech may also lead to the creation of new AI-driven products and services, addressing specific industry needs and pain points. With Sarvam's elevated status, the company may face increased scrutiny and expectations, particularly regarding security and the responsible development of AI systems. As the AI landscape continues to evolve, Sarvam's journey will be closely watched, and its advancements may have significant implications for the broader ecosystem, including companies like Respond.io, Salesforce, and SpaceX, which are also exploring AI-driven solutions.

[09:25] Respond.io Raises $62.5M

Respond.io's $62.5 million funding raise is a notable development in the AI agent-powered messaging app space. The company's use of AI agents to handle customer inquiries has enabled it to charge per conversation, rather than per seat, making it an attractive option for businesses looking to streamline their customer support workflows. Respond.io's architecture utilizes API integration with local coding models, enabling low-latency inference and configurable runtime behavior. The company's SDK allows for easy deployment and configuration of AI agents, and its runtime environment is designed to optimize performance and minimize latency. With this funding, Respond.io plans to acquire companies in North America and Europe, expanding its reach and solidifying its position in the market. This move is likely to have a significant impact on the customer support industry, as AI-powered messaging apps become increasingly prevalent. As Respond.io continues to grow and expand, it will be important to watch how it navigates the complexities of AI labor risk and ensures the security and integrity of its AI agents.

[10:28] Salesforce Acquires Fin AI Platform

Salesforce's acquisition of Fin, announced on June 15, 2026, marks a significant move in the AI customer service space. The $3.6B deal will see Fin's team and technology integrated into Salesforce's Agentforce platform, which allows businesses to build custom AI agents to automate tasks. The integration of Fin's API and architecture with Agentforce will enable enhanced automation capabilities, leveraging runtime behavior and protocol details to facilitate seamless interactions. This acquisition enables businesses to provide more effective customer service and streamline their operations, and it will be interesting to see how the combined capabilities of Fin and Agentforce play out in the market. With the recent release of Codex rust-v0.140.0, the new stable agent-stack release, developers can expect even more robust AI agent capabilities. As the AI labor risk landscape continues to evolve, this acquisition highlights the growing importance of AI-driven customer service and automation. The acquisition is also likely to have implications for other players in the space, such as SpaceX and its Cursor/Anysphere initiatives, as well as Respond.io and Sarvam, and will likely influence the development of local coding models and Rio/Nex-N2 provenance.

[11:37] AI Layoff Wave Intensifies

The recent escalation of the AI layoff wave has significant implications for the industry and the workforce. With the release of Codex rust-v0.140.0, the stable agent-stack version, developers can leverage the improved performance and efficiency of local coding models. However, this also means that the demand for human labor in certain sectors may decrease, exacerbating the layoff trend. The API and SDK configurations of these models play a crucial role in determining their inference latency and overall performance. Furthermore, the architecture of these models and their deployment strategies can significantly impact the labor market. As the industry continues to evolve, it is essential to consider the potential consequences of automated workflows and the need for responsible AI development practices. The contrast between the wealth accumulated by AI insiders and the layoffs affecting tens of thousands of workers highlights the risks and challenges associated with the AI labor market. The adoption of Rio/Nex-N2 provenance and Apple Foundation Models may also influence the market, and developers should be aware of these factors when designing and deploying AI-powered systems. What to watch next is how the industry responds to these challenges and whether it can find a balance between innovation and social responsibility.

[12:52] OpenAI Partner Network Launched

OpenAI has launched the Partner Network, a $150M initiative aimed at accelerating enterprise AI adoption and deployment. This network will provide partners with access to resources, support, and funding to develop and deploy AI-powered solutions. The Partner Network will leverage OpenAI's API and SDK to enable partners to build and deploy AI solutions, with a focus on security, scalability, and collaboration. With the Partner Network, OpenAI is investing in the growth and success of its partners, providing them with the tools and expertise needed to succeed in the AI landscape. This launch is significant, as it will enable businesses to transform and grow through the adoption of AI. The Partner Network will also provide a platform for partners to collaborate and innovate, driving the development of new AI-powered solutions. What's more, the network's focus on enterprise AI adoption and deployment will help to address some of the key challenges businesses face when implementing AI, such as integration, deployment, and maintenance. As the AI landscape continues to evolve, the OpenAI Partner Network is well-positioned to play a key role in shaping the future of enterprise AI. With Codex rust-v0.140.0 as the new stable agent-stack release, the possibilities for AI-powered solutions are endless, and the OpenAI Partner Network is poised to help businesses capitalize on these opportunities.

[14:14] Practical queue

From today's stories: For builders, this shifts what the stack can rely on by default. What this means for builders and workflows is that they need to consider the potential risks and benefits of merging existing models, as well as the implications for model provenance and ownership. For builders, this shifts what the stack can rely on by default. What this means for builders is that they will need to consider the identity and security implications of AI agents in their workflows, and NewCore's solution provides a potential approach to addressing these concerns. What this means for builders is that they can expect new opportunities for AI-powered workflows and applications, particularly in areas like space exploration and development. What this means for builders is that Sarvam's newfound resources may lead to advancements in AI research and development, potentially creating new opportunities for collaboration and innovation. What this means for builders is that they can leverage AI agent-powered messaging apps like Respond.io to streamline customer support workflows. What this means for builders is that they can expect improved AI agent capabilities and more efficient automation workflows. What this means for builders is that they must be aware of the potential consequences of their work on the labor market and consider the ethical implications of automated workflows. What this means for builders is that they will have access to more resources and support to develop and deploy AI solutions.

Chapters

• 00:00 — Intro: Agent Stack Release Readout: OpenAI Codex rust-v0.140.0 / Rio's Homegrown LLM / Apple Foundation Models
• 02:00 — Agent Stack Release Readout: OpenAI Codex rust-v0.140.0
• 03:03 — Rio's Homegrown LLM
• 04:27 — Apple Foundation Models
• 05:45 — NewCore Launches AI Agent Identities
• 07:20 — SpaceX Acquires Cursor for $60B
• 08:21 — Sarvam Reaches Unicorn Status
• 09:25 — Respond.io Raises $62.5M
• 10:28 — Salesforce Acquires Fin AI Platform
• 11:37 — AI Layoff Wave Intensifies
• 12:52 — OpenAI Partner Network Launched
• 14:14 — Practical queue

Primary Links

• OpenAI Codex rust-v0.140.0 release: https://github.com/openai/codex/releases/tag/rust-v0.140.0
• Ask HN: Has anyone replaced Claude/GPT with a local model for daily co: https://news.ycombinator.com/item?id=48542100
• Rio de Janeiro's "homegrown" LLM appears to be a merge of an existing : https://github.com/nex-agi/Nex-N2/issues/4
• Apple Foundation Models: https://platform.claude.com/docs/en/cli-sdks-libraries/libraries/apple-foundation-models
• As AI agents become employees, NewCore emerges with $66M to give them : https://techcrunch.com/2026/06/15/ai-agents-are-becoming-employees-newcore-emerges-with-66m-to-give-them-identities/
• SpaceX to acquire Cursor for $60B in stock, days after blockbuster IPO: https://techcrunch.com/2026/06/16/spacex-to-acquire-cursor-for-60b-in-stock-days-after-blockbuster-ipo/
• Sarvam becomes India’s newest AI unicorn with $234 million funding rou: https://techcrunch.com/2026/06/15/sarvam-becomes-indias-newest-ai-unicorn-with-234-million-funding-round-led-by-hcltech/
• Malaysia’s AI agent-powered messaging app Respond.io raises $62.5M, ey: https://techcrunch.com/2026/06/15/malaysias-respond-io-raises-62-5m-eyes-acquisitions-in-north-america-and-europe/
• Salesforce acquires AI customer service platform Fin for $3.6B: https://techcrunch.com/2026/06/15/salesforce-acquires-ai-customer-service-platform-fin-for-3-6b/
• The AI layoff wave is becoming a powder keg: https://techcrunch.com/2026/06/15/the-ai-layoff-wave-is-becoming-a-powder-keg/
• Introducing the OpenAI Partner Network: https://openai.com/index/introducing-openai-partner-network
• PrefectHQ/fastmcp repo: https://github.com/PrefectHQ/fastmcp
• microsoft/mcp-for-beginners repo: https://github.com/microsoft/mcp-for-beginners
• CoplayDev/unity-mcp repo: https://github.com/CoplayDev/unity-mcp
• Meta reportedly moves to unwind $2B Manus deal after Beijing’s demand: https://techcrunch.com/2026/06/13/meta-reportedly-moves-to-unwind-2b-manus-deal-after-beijings-demand/
• OpenAI Losses Increased Nearly 8X in 2025, with Spending Hitting $34B: https://www.wheresyoured.at/exclusive-openai-financials/
• Publishing WASM wheels to PyPI for use with Pyodide: https://simonwillison.net/2026/Jun/13/publishing-wasm-wheels/#atom-entries
• Ollama v0.30.8: https://github.com/ollama/ollama/releases/tag/v0.30.8

Release Coverage Check

• OpenClaw — Latest stable verified: v2026.6.6, published 2026-06-12T11:04:42Z. Recent episode version tags detected: v2026.6.5-beta.6, v2026.6.6, v2026.6.7-beta.1, v2026.6.8-beta.1. No new stable release this cycle.
• Hermes Agent — Latest stable verified: v2026.6.5, published 2026-06-06T00:55:58Z. Recent episode version tags detected: v0.15.2, v0.16.0, v2026.5.29.2, v2026.6.5. No new stable release this cycle.
• OpenAI Codex — Latest stable verified: rust-v0.140.0, published 2026-06-15T21:06:37Z. Recent episode version tags detected: rust-v0.135.0, rust-v0.137.0, rust-v0.138.0, rust-v0.139.0. Selected missing version(s): rust-v0.140.0.
• Claude Code CLI — Latest stable verified: 2.1.153, published (date not in registry window). Recent episode version tags detected: 2.1.168, 2.1.169, latest, stable. No new stable release this cycle.
• Antigravity CLI — Continuous delivery model; no discrete release tags verified this cycle (latest build as of 2026-06-16). Recent episode version tags detected: none on record.

Harness Version Reference

• OpenClaw — v2026.6.6 (stable) / v2026.6.8-beta.2 (prerelease)
• Hermes Agent — v2026.6.5
• OpenAI Codex — rust-v0.140.0
• Claude Code CLI — 2.1.153
• Antigravity CLI — Continuous delivery (no tagged release verified this cycle)