Local Mental Health Assessment Report Generator

Problem context

When mental-health support generates lots of unstructured notes, manual review is slow and can miss emotion, cognition, behavior, and risk cues. mental-health-monitor targets this workflow by using a local Ollama model to turn personal entries, group chats, questionnaire text, or CSV data into a reviewable structured report.

How it works

The skill centers on local inference and report generation. Its main capabilities include:
- Structured reports: supports individual mode and group mode; in group chats, higher-risk members can receive detailed entries while others are summarized.
- Scale scoring: automatically estimates PHQ-9 depression screening and GAD-7 anxiety screening from the content.
- Risk assessment: applies C-SSRS to grade suicide/self-harm risk and include crisis-response resources.
- Intervention guidance: references evidence-based techniques such as CBT, DBT, and ACT, with professional referral prompts.
A typical flow is to confirm a local Ollama service, select a model, provide text or a file, and receive the generated report. Connection settings live in references/config.md, while prompt and scale guidance live in references/system_prompt.md.

Boundaries

This tool is not a diagnostic tool and cannot replace psychological counseling or psychiatric care. Sensitive data is processed by the local Ollama endpoint by default, but output quality depends on input quality and model capability. In emergencies involving suicide or self-harm, contact local emergency services or a crisis hotline immediately.