Open-source safety infrastructure for the internet. Plugins, APIs, and frameworks — all designed to be dropped into any platform with minimal friction.
Every repo is open for contributions. Beginner-friendly issues are tagged. Good first PR waiting for you.
A lightweight browser plugin that flags harmful content patterns in real time. Configurable by age group. Designed to feel native, not bolted on.
REST API for detecting grooming language patterns, escalation behavior, and contact solicitation. Free tier for nonprofits and researchers.
A game framework and curriculum for teaching digital literacy. Think "Oregon Trail" energy, but the wagon is your privacy settings.
A unified reporting layer that aggregates safety signals across platforms. Non-surveillance: activity summaries, not keystroke logs.
An open standard for evaluating platform safety policies. Automated scoring based on public policy documents. No more self-grading.
A wrapper layer for LLM APIs that intercepts and flags age-inappropriate generation attempts before they reach a child user.
Five minutes from clone to first detection. We kept the setup boring on purpose — so the work can be interesting.
# Clone the filter plugin git clone https://github.com/The-Cocoon-Collective/cocoon-filter cd cocoon-filter && npm install # Configure for your platform cp config.example.json config.local.json # Run the dev build npm run dev # Your first safety check import { CocoonFilter } from '@cocoon-collective/filter' const result = await CocoonFilter.check({ text: "message content here", profile: "under-13", flags: ["contact-solicitation", "age-mismatch"] }) # result.safe → true/false result.signals → [...]
No experience with safety tech required — just a willingness to make something good. Every language background welcome. Every skill level needed.