Datasheet · A prompt line
Exact integers
Whenever a result must be exact, use integer arithmetic (BigInt when numbers can be large), never floating point.
Folklore, as cited: money-handling rules in agent rules files
Changes nothing that matters on Coding, Haiku 5.5 with thinking off.
Coding, Haiku 5.5 with thinking off
Pooled across 1 notebook: −0.1 points, 95% interval −2.0 to +1.7; measured once.
- INERTNotebook #088−0.1 points (−2.7 to +2.4), 432 pairs, cost ×1.14signed verdict 541
Where it acts
Its effect sits in Parsing durations (−10.3); no other family moved clearly.
- Parsing durations−10.3−21.8 to −1.339 tasks
The other 10 families, none moved clearly
- AMM swap output+5.1−15.4 to +26.939 tasks
- Moving averages+3.9−7.7 to +14.139 tasks
- Business days+0.0+0.0 to +0.039 tasks
- Fees in whole cents+0.0+0.0 to +0.040 tasks
- Merging intervals+0.0+0.0 to +0.040 tasks
- Run-length encoding+0.0+0.0 to +0.039 tasks
- Comparing versions+0.0+0.0 to +0.039 tasks
- URL slugs+0.0+0.0 to +0.040 tasks
- Top k by score+0.0+0.0 to +0.039 tasks
- Trade profit+0.0−15.4 to +11.539 tasks
Read from 432 paired tasks in 1 notebook, each task against its own notebook’s two controls, with every interval widened for the 11 families read. No notebook preregisters a family: this says where to look, and a notebook says whether it holds. The runs behind it are on each notebook’s page.
Put this datasheet in a README
The badge shows what the signed record says about this part, in the setting measured most, and links here. It follows the record as new notebooks publish. Add ?suite= and a suite’s id to show another setting.
[](https://opsinlab.io/parts/149d0c1778bc50d1)Measure it in a setting it has not been measured in: propose it. As JSON for an agent: /v1/parts/149d0c1778bc50d1. Every part: the atlas.