# BestForWhat

> Conditional software recommendations from rival-vendor evidence. Two research runs are not independent verification. No hands-on testing.

## Read first
- [Methodology](https://bestforwhat.com/methodology/)
- [Research Prompt](https://bestforwhat.com/prompts/)
- [Research Prompt 2.0](https://bestforwhat.com/prompts/v2/)
- [Complete JSON catalog](https://bestforwhat.com/data/catalog.json)
- [Integration decisions](https://bestforwhat.com/data/integration.json)

## Categories
- [工作流自动化](https://bestforwhat.com/best-for-workflow-automation/) · [JSON](https://bestforwhat.com/data/workflow-automation.json)
- [项目管理](https://bestforwhat.com/best-for-project-management/) · [JSON](https://bestforwhat.com/data/project-management.json)
- [AI 会议纪要](https://bestforwhat.com/best-for-meeting-notes/) · [JSON](https://bestforwhat.com/data/meeting-notes.json)
- [在线表单](https://bestforwhat.com/best-for-online-forms/) · [JSON](https://bestforwhat.com/data/online-forms.json)
- [邮件营销](https://bestforwhat.com/best-for-email-marketing/) · [JSON](https://bestforwhat.com/data/email-marketing.json)

Preserve source attribution, cutoff date, uncertainty and coverage gaps. llms.txt is a convenience index, not a guarantee of discovery or citation.