[PR #32444] improve frontend-code-review skill description and review process #33747

Open
opened 2026-02-21 20:53:49 -05:00 by yindo · 0 comments
Owner

Original Pull Request: https://github.com/langgenius/dify/pull/32444

State: open
Merged: No


hey @langgenius, thanks for building dify in the open and making your agent skills available to the community. Kudos on soon hitting 130k stars! I just starred it.

was running your skills through some evals and noticed a few things on frontend-code-review that were pretty quick to improve (moving from 68% to 100% agent performance):

  • expanded capabilities description + trigger terms like code review, PR review, component review so agents can reliably match user requests to this skill

  • added urgency classification criteria with inline examples so agents know when to flag urgent vs suggestion

  • added 3 concrete violation examples directly in the review process steps (classnames without cn(), inline object props, workflowStore misuse)

these were easy changes to bring the skill in line with what performs well against Anthropic's best practices. honest disclosure, I work at a company where we build tooling around this. not a pitch, just fixes that were straightforward to make.

you've got 4 skills here, if you want to do it yourself, evals are free and open to run: link otherwise happy to make the improvements for you.

**Original Pull Request:** https://github.com/langgenius/dify/pull/32444 **State:** open **Merged:** No --- hey @langgenius, thanks for building dify in the open and making your agent skills available to the community. Kudos on soon hitting 130k stars! I just starred it. was running your skills through some evals and noticed a few things on frontend-code-review that were pretty quick to improve (moving from `68%` to `100%` agent performance): - expanded capabilities description + trigger terms like _code review, PR review, component review_ so agents can reliably match user requests to this skill - added urgency classification criteria with inline examples so agents know when to flag _urgent_ vs _suggestion_ - added `3` concrete violation examples directly in the review process steps (_classnames without cn(), inline object props, workflowStore misuse_) these were easy changes to bring the skill in line with what performs well against **Anthropic's best practices**. honest disclosure, I work at a company where we build tooling around this. not a pitch, just fixes that were straightforward to make. you've got `4` skills here, if you want to do it yourself, evals are free and open to run: [link](https://tessl.io/registry/skills/github/langgenius/dify) otherwise happy to make the improvements for you.
yindo added the pull-request label 2026-02-21 20:53:49 -05:00
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: langgenius/dify#33747