Hoplite (YC S26) Launches Cloud Deployment for Coding Agents with QA Previews
Hoplite (hoplite.sh) is a new service from YC S26 that deploys coding agents in the cloud with a focus on easily QA-ing features. It ports your local setup—sessions, memories, MCP servers—and gets your projects ready to run in the cloud. The founders, Bence and Ryan, built a custom agent harness instead of using off-the-shelf solutions like Codex or Claude Code, to have independence and test new features without waiting for Anthropic or OpenAI.
Key Features
- Custom agent harness: Built in-house, giving full control and flexibility to experiment with new agent behavior.
- Cloud infrastructure: Hosted on AWS, with Temporal for durable workflows, Modal for sandboxes, and Planetscale for the database. They emphasize reliability and security as agents become tier-0 infrastructure.
- Onboarding: Transfers your local sessions, memories, and MCP servers to the cloud, making setup seamless.
- Previews for QA: The team is optimizing previews to let developers visually verify new user flows, UI, API responses, and CLI behavior across platforms like Windows, while running hundreds of agents concurrently.
Pricing and Trial
You can try Hoplite for free with the code HACKERNEWS, which includes $100 in free credits. You can also connect your Codex subscription to use OpenAI models through Hoplite. More details at hoplite.sh/pricing.
Who It's For
Developers who want to run AI coding agents at scale in the cloud and need an efficient way to QA the results, especially those who expect to review product output rather than code as models improve.
📖 Read the full source: HN LLM Tools
👀 See Also
Managing Multiple AI Agent Tasks with Kanban Boards
A developer shares their experience running multiple Claude AI agents in terminal tabs and identifies three key workflow challenges: lack of progress visibility, context loss when switching between tasks, and rate limit interruptions. Their solution involves treating AI tasks like work items on a Kanban board.
The Human Creativity Benchmark: Separating Convergence from Divergence in AI Creative Evaluation
Contra Labs introduces the Human Creativity Benchmark (HCB), a framework that distinguishes objectively verifiable criteria (e.g., prompt adherence) from subjective taste (e.g., visual appeal) in evaluating generative AI for creative work. The benchmark reveals that no current model is reliably both correct and steerable, addressing mode collapse and the need for differentiated output.
Cull: Open-Source Dataset Curation Engine for AI Image Pipelines
Cull scrapes images from 340+ sources including Civitai, X/Twitter, Reddit, Discord, and booru sites, classifies them with a vision-language model via local LM Studio or Groq, and sorts into category folders with SD prompts and audit records.
Optio: Orchestrating AI Coding Agents in Kubernetes from Ticket to PR
Optio is an open-source orchestration system that turns tickets into merged pull requests using AI coding agents like Claude Code or Codex. It handles the full lifecycle in isolated Kubernetes pods with a feedback loop that auto-resumes agents on CI failures or review feedback.