Clamshell Keeps Mac Agents Running Lid-Closed

// 1h agoPRODUCT LAUNCH

Clamshell Keeps Mac Agents Running Lid-Closed

Clamshell lets Apple Silicon MacBooks keep agents, builds, downloads, and terminal sessions running after the lid closes, without an external monitor, dock, sudo, or admin setup. It supports macOS 14+ and works on AC power or battery.

// ANALYSIS

Clamshell solves a narrow but genuinely painful failure mode for developers running local agents and long jobs. Its simplicity is compelling, though sustained workloads still demand attention to battery life, airflow, and thermals.

–Option-close activation keeps the workflow quick and unobtrusive
–The display fully sleeps while network access and processes remain active
–No privileged helper, driver, kernel extension, or external display is required
–A $9.99 one-time purchase with a seven-day trial is refreshingly straightforward
–Battery and thermal safeguards matter for unattended overnight runs, especially on fanless MacBook Air models

// TAGS

clamshelldevtoolautomationagentcoding-agentlocal-first

DISCOVERED

1h ago

2026-08-15

PUBLISHED

7h ago

2026-08-15

RELEVANCE

6/ 10

AUTHOR

Matyas

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

OPEN SOURCE14m ago

AutoDesign evolves frozen LLM harnesses

AutoDesign is an open-source design agent that recursively improves its prompts, tools, orchestration, validation, and rendering logic around fixed language models. Its DesignHarness converts research papers into editable posters, slides, webpages, and narrated videos through iterative generation and critique.

TUTORIAL28m ago

OpenCode Go Powers $10 DeepSeek Stack

An X post showcases a low-cost coding setup combining OpenCode Go’s $10/month access to open models with the open-source DeepSeek Harness. The harness adds protocol handling, CLI, MCP, and agent integrations for DeepSeek V4.

OPEN SOURCE2h ago

Soup CLI Launches Local 8B Fine-Tuning

Soup CLI simplifies LLM fine-tuning into one YAML configuration and command, with layer streaming that reportedly trains Llama 3.1 8B using 3.32 GB of VRAM. It supports methods including SFT, DPO, GRPO, and KTO, plus evaluation and model export.