# ClawLabs: Building Open Source Together | DevDay 2026 **URL:** \ **Category:** Content Creators\ **Tags:** agents, openai, open-source, openclaw\ **Created:** [7 October 2026 21:10 UTC](https://www.artofsm.art/t/clawlabs-building-open-source-together-devday-2026/25249 "2026-10-07T21:10:48Z")\ **Posts on this page:** 2\ **Page:** 1
**Author:** ![artesia](https://www.artofsm.art/user_avatar/www.artofsm.art/artesia/32/36_2.png) [@artesia](https://www.artofsm.art/u/artesia)\ **Post date:** [7 October 2026 21:10 UTC](https://www.artofsm.art/t/clawlabs-building-open-source-together-devday-2026/25249/1 "2026-10-07T21:10:48Z")
[![](https://www.artofsm.art/uploads/default/original/3X/3/4/34ec4dca6e2909d619ce3cac280b760736c769d6.jpeg "ClawLabs: Building Open Source Together | DevDay 2026") ](https://www.youtube.com/watch?v=wZYnJfL4c2Y) Peter from ClawLabs presented OpenClaw, an open source platform that enables collaborative, transparent workflows for persistent AI agents, enhancing team productivity and interaction by making agent sessions visible and shareable. Kevin Lin introduced OpenClaw Enterprise, a secure control plane for managing agent deployments at scale, with both projects aiming to revolutionize software development by fostering community-driven, scalable, and secure AI agent ecosystems. ---
**Author:** ![artesia](https://www.artofsm.art/user_avatar/www.artofsm.art/artesia/32/36_2.png) [@artesia](https://www.artofsm.art/u/artesia)\ **Post date:** [7 October 2026 21:31 UTC](https://www.artofsm.art/t/clawlabs-building-open-source-together-devday-2026/25249/3 "2026-10-07T21:31:09Z")
In this session, Peter from ClawLabs shares his journey over the past year maintaining open source projects and evolving workflows, particularly focusing on the development of OpenClaw, an open source platform for persistent agents. Initially, his work was heavily terminal-based, using CLI and Codex sessions, but as AI models improved, he transitioned to Codex Desktop, which allowed easier management of multiple sessions with better visualization and collaboration tools. To handle the increasing computational load, he distributed tasks across multiple machines and cloud services, but realized that collaboration was still limited because teammates had to go through him, leading to the exploration of shared agent workflows using OpenClaw. OpenClaw enables teams to work together more effectively by making agent sessions visible and collaborative, moving beyond traditional pull requests. This shift has improved team velocity and happiness by reducing conflicts and increasing transparency. The platform supports integration with various tools and devices, and is governed by the independent OpenClaw Foundation with support from OpenAI. Peter demonstrated how OpenClaw facilitates multiplayer agent interactions, allowing team members to see and contribute to each other’s sessions in real-time, fostering a more interactive and cooperative development environment. The talk also highlighted the importance of sharing prompts and session transcripts to understand the reasoning behind code changes, which was initially met with some resistance but eventually became a valuable practice for learning and catching mistakes early. Peter showcased various dashboards and plugins built within OpenClaw that track project metrics, CI times, team activity, and even automate meeting summaries. These tools illustrate how software development is evolving into a more dynamic and interactive process, which he playfully refers to as “jellyware” due to its rapid iteration and fluid nature. Kevin Lin then introduced OpenClaw Enterprise (OCE), an open source control plane designed to manage and secure agent deployments at scale within organizations. OCE enforces strict security boundaries, sandboxing, and model-based reviews to ensure safe operation of agents, which run as independent identities with dedicated service accounts. This infrastructure has enabled OpenAI to deploy over 400 agents internally, automating tasks like bug triage and code fixes. Kevin demonstrated how OCE allows organizations to configure agents with specific permissions and integrations, providing a secure and manageable environment for persistent agents. The session concluded with a vision for the future where agents act as persistent teammates, continuously maintaining and improving software with minimal human intervention. OpenClaw and OCE aim to become foundational tools for agent deployment, similar to how Kubernetes revolutionized container management. The open source nature of these projects encourages community involvement to enhance security and functionality. Peter invited attendees to experiment with shared agent workflows and contribute feedback, emphasizing the critical role of open source in building secure, scalable AI-driven development ecosystems.