I built a micro script to monitor OpenRouter and notify me on Slack (github.com)

A small open-source Bash monitor watches the OpenRouter models API (https://openrouter.ai/api/v1/models) and posts Slack alerts whenever new models appear. The script (monitor_openrouter_models.sh) fetches the model list, compares model IDs against a local state file (.openrouter_models_last.json), and — when it finds new entries — resolves each model’s display name and slug, then sends a Slack message (via an Incoming Webhook set in .env) with the model name, ID, and a direct link (https://openrouter.ai/models/{slug}). The first run simply initializes the baseline so you don’t get a flood of notifications. The repo includes .env.example, ignores local state and secret files, and relies on standard CLI tools (bash, curl, jq, comm/coreutils). This approach is significant because it gives ML engineers and researchers a lightweight, self-hosted way to get real-time visibility into new model releases without relying on heavy tooling or third-party trackers. It’s easy to deploy (chmod +x, set SLACK_WEBHOOK_URL, run once, add a cron job — example shown for every 15 minutes), and is extendable: you can add filters, alternative notification channels, rate-limit handling, or richer metadata parsing. Because it polls a public API, consider polling frequency and error handling to avoid rate limits and ensure reliable alerts.