> Markdown version of [/videos/1897-agents-version-control-and-bunnies-daniel-siegl-david-payr](https://www.wearedevelopers.com/videos/1897-agents-version-control-and-bunnies-daniel-siegl-david-payr). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Agents, Version Control and Bunnies - Daniel Siegl & David Payr AI agents need architectural guardrails. Daniel Siegl and David Payr reveal how to safely sandbox them using Git Worktrees. Plus, discover zero-dependency browser media processing with MediaBunny. - **Speakers:** [Chris Heilmann](https://www.wearedevelopers.com/@chris-heilmann), [Daniel Cranney](https://www.wearedevelopers.com/@daniel-cranney) - **Event:** WeAreDevelopers LIVE - **Published:** June 3, 2026 - **Duration:** 1:04:56 - **URL:** https://www.wearedevelopers.com/videos/1897-agents-version-control-and-bunnies-daniel-siegl-david-payr ## Summary The modern developer ecosystem is rapidly evolving to incorporate AI generation and complex browser APIs, requiring both tighter architectural guardrails and highly precise local tooling. Addressing the burden of client-side media processing, the MediaBunny TypeScript library replaces heavy-handed, server-side FFmpeg or bulky WebAssembly dependencies with direct hardware integration. By leveraging the native WebCodecs API, developers gain frame-perfect video seeking and file generation directly in the browser. This zero-dependency, tree-shakeable approach allows engineers to render HTML canvases to MP4s, extract cross-browser audio formats, and validate client-side video metadata without inflating bundle sizes or sacrificing local hardware acceleration. Accelerating development velocity through AI coding agents necessitates matching that speed with robust repository safety mechanisms. To prevent AI from autonomously corrupting environments, tools like Syntevo's SmartGit leverage the historically underutilized "Git Worktree" feature as a protective sandbox. Restricting LLM agents to isolated workspaces allows them to rapidly iterate, refactor, and generate files without contaminating the primary feature branch. Human engineers can experiment with varied codebase solutions asynchronously, thoroughly review the AI's output, and selectively cherry-pick successful commits. This methodology maintains a pristine commit history while elevating developers to the role of vigilant architectural reviewers. Applying these concepts reinforces the fundamental distinction between haphazard prompt engineering and disciplined software infrastructure. As developers explore agentic architecture in environments like the upcoming AWS hackathons, maintaining explicit audit trails becomes critical for security and compliance. Treating AI models as high-speed typing accelerators means enforcing strict verifications—such as SSH commit signing to prevent identity spoofing via stolen tokens, and attaching LLM-generated rationale via Git Notes. Securing the workflow guarantees that modern code generation remains secure, thoroughly documented, and maintainable for long-term production lifecycles. **Keywords:** typescript media processing, webcodecs video encoding, browser-based media manipulation, hardware-accelerated browser processing, ffmpeg webassembly alternatives, agentic development workflow, git worktree sandboxing, smartgit version control, ssh commit signing security, llm audit trails, zero-dependency bundle optimization, client-side video metadata validation, git notes integration, ai agent repository guardrails ## Chapters 1. **Building AI applications at the AWS football hackathon** (05:13) — Developers can explore new APIs and build artificial intelligence solutions at the upcoming WeAreDevelopers hackathon. 1. **Processing media natively in the browser with MediaBunny** (13:22) — The open-source MediaBunny library leverages the WebCodecs API for efficient system media writing without external servers. 1. **Reviewing recent technology and AI tool headlines** (30:33) — A review of current events highlights the real-world operational costs and risks associated with emerging coding tools. 1. **Isolating agent-generated code with Git worktrees** (37:26) — Isolating autonomous code generation in separate Git worktrees allows developers to safely review changes before integration. 1. **Managing AI code branches with SmartGit worktrees** (43:54) — Visual interfaces like SmartGit streamline the process of organizing, reviewing, and cherry-picking commits generated by autonomous systems. 1. **Preventing impersonation with Git notes and signed commits** (48:03) — Using Git notes and SSH key commit signing secures repositories against malicious token deployment and developer impersonation. 1. **Preserving engineering fundamentals in agentic development** (56:01) — Human accountability, robust requirements gathering, and deliberate system design remain essential practices when deploying automated coding agents. ## Related Moments - [Security integration and AI skepticism in developer tooling](https://www.wearedevelopers.com/videos/1830-wearedevelopers-live-speculaitions) (from "WeAreDevelopers LIVE - SpeculAItions") - [Shifting developer workloads and realistic AI productivity gains](https://www.wearedevelopers.com/videos/1830-wearedevelopers-live-speculaitions) (from "WeAreDevelopers LIVE - SpeculAItions") - [Bringing JavaScript developers into web artificial intelligence](https://www.wearedevelopers.com/videos/1896-how-web-ai-can-power-the-agentic-web-jason-mayes-google) (from "How Web AI Can Power the Agentic Web - Jason Mayes (Google)") - [Enforcing developer accountability when leveraging AI programming agents](https://www.wearedevelopers.com/videos/1752-what-ai-can-learn-from-version-control-daniel-siegl-syntevo) (from "What AI Can Learn from Version Control - Daniel Siegl (Syntevo)") - [Building agentic artificial intelligence applications using Node.js](https://www.wearedevelopers.com/videos/1512-building-ai-applications-with-langchain-and-node-js) (from "Building AI Applications with LangChain and Node.js") - [Recreating React components using AI and dev tools](https://www.wearedevelopers.com/videos/1786-wearedevelopers-live-ai-freelancing-keeping-up-with-tech-and-more) (from "WeAreDevelopers LIVE – AI, Freelancing, Keeping Up with Tech and More") ## Related Articles - [Dev Digest 132 - Binging WADFlix?](https://www.wearedevelopers.com/magazine/473-dev-digest-132-binging-wadflix) - [Dev Digest 108 - Git off my cloud!](https://www.wearedevelopers.com/magazine/407-dev-digest-108-git-off-my-cloud) - [What is Agentic Programming and Why Should Developers Care?](https://www.wearedevelopers.com/magazine/625-what-is-agentic-programming-and-why-should-developers-care) - [Dev Digest 137 - AI'm not sure about this](https://www.wearedevelopers.com/magazine/485-dev-digest-137-ai-m-not-sure-about-this) ## Related Jobs - [AI Software Engineer (Germany)](https://www.wearedevelopers.com/jobs/48317-ai-software-engineer-germany) at **Sunhat** - [Principal Product Manager, Agent Platform](https://www.wearedevelopers.com/jobs/ext/277541-principal-product-manager-agent-platform) at **GitHub** - [Staff Developer Advocate, GitHub Security Lab](https://www.wearedevelopers.com/jobs/ext/1921051-staff-developer-advocate-github-security-lab) at **GitHub** - [AI Full Stack Engineer](https://www.wearedevelopers.com/jobs/ext/1354435-ai-full-stack-engineer) at **Almedia** - [Staff Software Engineer, Copilot Experiences](https://www.wearedevelopers.com/jobs/ext/164361-staff-software-engineer-copilot-experiences) at **GitHub** - [Senior AI Agent Software Engineer (Go, Python) (m/f/x)](https://www.wearedevelopers.com/jobs/48277-senior-ai-agent-software-engineer-go-python-m-f-x) at **Dynatrace**