by UditAkhourii · Codex Skill · ★ 896
ADHD — a skill for agents An architectural fix for premature convergence in autoregressive reasoning. 📄 Preprint: ADHD: Parallel Divergent Ideation for Coding Agents 👤 Author: Udit Akhouri — @akhouriudit · LinkedIn · researchudit@gmail.com Linear Chain-of-Thought gets trapped in local minima: each generated token conditions the next, so the model anchors on whatever it said first. Tree-of-Thought widens the search but still walks a single shared context, so anchoring persists across branches.
| Stars | 896 |
| Forks | 47 |
| Language | TypeScript |
| Category | Codex Skill |
| License | MIT |
| Quality Score | 66.6643895579224/100 |
| Open Issues | 10 |
| Last Updated | 2026-06-30 |
| Created | 2026-05-25 |
| Platforms | claude-code, codex, node |
| Est. Tokens | ~19k |
These tools work well together with adhd for enhanced workflows:
Explore other popular codex skill tools:
adhd is ADHD — a skill for coding agents. Tree-of-thought with pruning, built on the Claude & Codex Agent SDK. Fans out parallel divergent thoughts under different cognitive frames, scores, prunes traps, deep. It is categorized as a Codex Skill with 896 GitHub stars.
adhd is primarily written in TypeScript. It covers topics such as adhd, agents, ai.
You can find installation instructions and usage details in the adhd GitHub repository at github.com/UditAkhourii/adhd. The project has 896 stars and 47 forks, indicating an active community.
adhd is released under the MIT license, making it free to use and modify according to the license terms.