Analysis updated 2026-05-18
Record a chronological timeline of what happened after an incident abroad
Generate a neutral English draft in the style of a police statement
Create separate drafts tailored for hospitals, universities, or insurers
Get a list of local support resources based on your jurisdiction
| liangchenwei666-ai/safeabroad-agent | adya84/ha-world-cup-2026 | afk-surf/safeclipper | |
|---|---|---|---|
| Stars | 16 | 16 | 16 |
| Language | Python | Python | Python |
| Setup difficulty | moderate | easy | moderate |
| Complexity | 2/5 | 2/5 | 3/5 |
| Audience | general | general | developer |
Figures from each repo's GitHub metadata at analysis time.
Works without an AI model but needs an LLM key configured for best results.
SafeAbroad Agent is a documentation tool designed to help Chinese-speaking international students record what happened after a frightening incident abroad, assault, street violence, threats, injuries, or property theft. The core problem it addresses is that after a traumatic event, a person may face language barriers, unfamiliar legal systems, insurance confusion, and psychological shock all at once, making it hard to organize memories and communicate clearly with police, hospitals, universities, or insurers. The tool works by conducting a trauma-informed intake conversation in Chinese, gently asking follow-up questions to fill in gaps and surface contradictions, then constructing a chronological timeline with explicit uncertainty markers, it never upgrades a fuzzy memory into a confident claim. From that timeline it can generate neutral English drafts in the style of a police statement, plus separate drafts tailored for medical providers, university administrators, police follow-up, insurance claims, and victim support services. It can also suggest local support resources based on jurisdiction. The system is built as a FastAPI backend with a static web interface. It can run with deterministic fallback logic even without an AI model, but connecting an LLM improves timeline extraction and wording quality. It supports Qwen, OpenAI, Anthropic, and any OpenAI-compatible endpoint, configured via a local .env file or an in-browser settings panel. Case state is modeled with Pydantic, and data contracts are documented as Zod schemas in a TypeScript reference layer. The project explicitly states it is not a replacement for emergency services, lawyers, doctors, or counsellors.
A trauma-informed intake tool that helps international students document an incident abroad and generates neutral drafts for police, hospitals, and insurers.
Mainly Python. The stack also includes Python, FastAPI, Pydantic.
License not stated in the available information.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly general.
This repo across BitVibe Labs
Verify against the repo before relying on details.