> ## Documentation Index
> Fetch the complete documentation index at: https://adhd.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Results

> Head-to-head numbers — ADHD vs a single-shot baseline across six open-ended engineering problems.

Mean scores across 6 open-ended engineering problems (0–10), ADHD vs a single-shot baseline at the same model, judged by an independent LLM with a skeptical-staff-engineer prompt, A/B order randomized per problem.

| Dimension          |     ADHD | Baseline |         Δ | Ratio |
| ------------------ | -------: | -------: | --------: | ----: |
| breadth            | **9.00** |     4.83 | **+4.17** |  1.9× |
| novelty            | **7.83** |     2.67 | **+5.17** |  2.9× |
| trap detection     | **9.50** |     1.83 | **+7.67** |  5.2× |
| actionability      | **9.50** |     6.50 | **+3.00** |  1.5× |
| builder usefulness | **7.67** |     6.83 | **+0.83** |  1.1× |

**ADHD wins 5 of 6 problems head-to-head.** The biggest gap is **trap detection** — single-shot baselines almost never name the seductive-but-broken ideas, while ADHD's separate critic pass routinely flags 15–20 of them with mechanistic reasons.

Full per-problem verdicts live in [`EVALS.md`](https://github.com/UditAkhourii/adhd/blob/main/EVALS.md); full transcripts in [`bench/results.json`](https://github.com/UditAkhourii/adhd/blob/main/bench/results.json). *Run date: 2026-05-25.*

## Independent external benchmarks

* [**A measured duel vs. single-shot**](https://miyagadget.page/en/blog/2026/06/03/adhd-coding-agent-skill-en/) by Shichinomiya — an independent blind-scored benchmark (2 problems, LLM-as-judge, A/B positions swapped). ADHD won both, with the biggest gains in novelty (4.5→9.0) and trap detection (5.0→9.0), at a real cost of \~2.3× time and \~1.9× output.
* [**Han plugin compatibility analysis**](https://github.com/testdouble/han/blob/adhd-swarm-research/docs/research/adhd-application-to-han.md) — an evidence-based review using Han's own `/research` skill: 11 sources, 8 validation rounds. Findings tracked openly as issues [#16](https://github.com/UditAkhourii/adhd/issues/16), [#17](https://github.com/UditAkhourii/adhd/issues/17), and [#18](https://github.com/UditAkhourii/adhd/issues/18).

## Known limitations

Stated plainly, because reviewers will find them anyway:

* **Same-model judging.** The judge is the same model family as the generator (familiarity bias). Cross-model judging is on the roadmap ([issue #6](https://github.com/UditAkhourii/adhd/issues/6)).
* **Small set.** Six problems, all engineering-shaped.
* **Scale gap.** Evals run at K=5 branches; the academic diversity literature measures at K=100. Bridging this is tracked in [issue #18](https://github.com/UditAkhourii/adhd/issues/18).
* **Human-in-the-loop applicability is unproven.** A controlled CHI 2025 study found no significant benefit from LLM problem-reframing with human designers. ADHD's LLM-to-LLM context differs, but this is honest counter-evidence, tracked in [issue #16](https://github.com/UditAkhourii/adhd/issues/16).

<Card title="Reproduce the numbers" icon="flask" href="/evals/methodology">
  Methodology, how the judge is prompted, and how to run the suite yourself.
</Card>