by Cranot · Agent Tool · ★ 124
AGI in md Your most expensive model produces shallow analysis because you're asking it to reason about problems instead of through them. A 332-word prompt fixes this. A prism is a markdown system prompt that acts as a cognitive program — it tells the model to do specific things in order: make a claim, attack it, build an improvement, watch what breaks, derive the trade-off that can't be escaped. This repo contains 58 prisms + 27 scan modes + the tooling to use them. What this is: A system for eliciting structural insight under controlled prompt programs.
| Stars | 124 |
| Forks | 16 |
| Language | Python |
| Category | Agent Tool |
| License | MIT |
| Quality Score | 68.4212158763378/100 |
| Last Updated | 2026-03-18 |
| Created | 2026-02-28 |
| Platforms | aws, claude-code, gemini, python |
| Est. Tokens | ~480k |
These tools work well together with agi-in-md for enhanced workflows:
Explore other popular agent tool tools:
agi-in-md is Cognitive prisms for structural analysis — any domain, any model. 1000+ experiments, 58 prisms, 27 scan modes. The cheapest model + right prism beats the most expensive without one. Cross-architecture. It is categorized as a Agent Tool with 124 GitHub stars.
agi-in-md is primarily written in Python. It covers topics such as ai-tools, claude, code-analysis.
You can find installation instructions and usage details in the agi-in-md GitHub repository at github.com/Cranot/agi-in-md. The project has 124 stars and 16 forks, indicating an active community.
agi-in-md is released under the MIT license, making it free to use and modify according to the license terms.