docs/listing/claude.md

# Claude Listing

Name: AI Psychiatry

Slug: `ai-psychiatry`

Marketplace: `ihabkhaled-ai`

Category: Productivity

Website: https://github.com/ihabkhaled/AI-Psychiatry

Support: https://github.com/ihabkhaled/AI-Psychiatry/blob/main/SUPPORT.md

Privacy: https://github.com/ihabkhaled/AI-Psychiatry/blob/main/PRIVACY.md

Terms: https://github.com/ihabkhaled/AI-Psychiatry/blob/main/TERMS.md

Logo: https://raw.githubusercontent.com/ihabkhaled/AI-Psychiatry/main/assets/ai-psychiatry-logo.png

Short description: Semantic executive control for reliable coding agents.

Long description: AI Psychiatry is a skills-only plugin that helps Claude Code preserve the primary objective, constrain causal nesting and WIP, detect strategy and scope laundering, reject false progress, completion, and blockers, protect memory and context, and balance underthinking against overthinking. It uses observable evidence contracts and bounded recovery procedures without adding an MCP server, analytics, or data collection. Human psychiatric terms are behavioral analogies only, never diagnoses.

Starter prompts:

- Install the AI Psychiatry framework while preserving this repository's existing instructions.
- Audit this task for semantic loopholes, hidden recursion, and strategy laundering.
- Balance underthinking and overthinking, then produce completion evidence.
- Never stop until the Definition of Done is proven (maximum autonomous execution, on request only).
- Use `/ai-psychiatry:all-the-medicine` - the one command - for the complete framework; it is also always on.
- Answer directly: terse, assertive, one non-looping question at most.

Hosted marketplace install:

```text
/plugin marketplace add ihabkhaled/AI-Psychiatry
/plugin install ai-psychiatry@ihabkhaled-ai
```

Community directory target: `ai-psychiatry@claude-community` after Anthropic review and approval.