Show HN: Oglama – an automated browser with built-in LLM and shareable modules (oglama.com)

Oglama is a new locally running automated browser/agent platform that combines deterministic web automation with an optional, small on-device LLM for edge cases like sentiment analysis or summarization. It’s built on Chromium, stores credentials, cookies and logs only on your machine, and isolates each agent in its own browser session (up to nine concurrent agents). Users can run, schedule and chain pre-built or custom modules (exportable as .oglama.yaml), create modules with beginner-level JavaScript, and share vetted modules via a community repository; public modules undergo manual review. The product emphasizes privacy (no telemetry, local data only), predictable pricing (flat subscription, no usage fees), and fully auditable source code for every module. For the AI/ML community this matters because it blends reliable, deterministic automation with selective local LLM augmentation—reducing the brittleness and privacy risks of always-on cloud LLMs while enabling higher-level tasks when needed. Technical implications include reproducible, auditable automation workflows suitable for large-scale scraping or complex UI flows, and a privacy-preserving pattern for integrating compact models into edge automation. It lowers the barrier to building and sharing reusable automation modules, but introduces a community-trust model for shared modules and reliance on locally available LLM capacity for advanced tasks.