shengjidaguai-china/goutoujunshi

What it solves

It helps users navigate complex romantic relationships—including dating, pursuit, conflicts, breakups, and reconciliation—by providing emotional support and actionable advice. Instead of generic suggestions, it analyzes specific chat logs, user narratives, and relationship dynamics to provide concrete next steps, such as specific messages to send or plans for a first date, while reducing emotional distress and overthinking.

How it works

Designed as a Codex Skill, the project uses a structured analysis pipeline: it first acknowledges the user's emotions, establishes profiles for the parties involved, separates facts from assumptions, and then retrieves relevant knowledge from a multidisciplinary library (covering relationship psychology, communication strategies, and legal safety). It can process text, chat screenshots, and exported chat files (via an optional ChatLab dependency) to analyze trends and provide tailored recommendations based on reciprocity, risk, and opportunity cost.

Who it’s for

Individuals seeking guidance on romantic interactions across all sexual orientations and gender identities, including those dealing with long-distance relationships, non-monogamous dynamics, or those struggling to interpret a partner's signals and decide whether to continue investing in a relationship.

Highlights

• Multidisciplinary Knowledge Base: Integrates 135 references across psychology, sociology, communication, and law to avoid "one-size-fits-all" formulas.
• Long-term Memory: Can maintain simplified profiles and progress tracking across different tasks with user consent.
• Context-Aware Analysis: Distinguishes between different data sources (screenshots vs. transcripts) and separates factual behavior from subjective speculation.
• Safety-First Approach: Includes specific modules for handling high-risk situations like domestic violence or fraud, prioritizing safety over dating tactics.