Two testing skills · both proven

systematic-debugging vs writing-guidelines

Both skills were run through Claude twice: once on a fixed brief without the skill, once with it. The briefs differ because each is written for the job the skill claims to do, so compare each skill against its own baseline, then decide which difference matters for your work.

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes

Category: Testing
Source: obra/superpowers
License: MIT
Proof model: claude-opus-5-5
Brief: “A Node.js API returns 500 for about 1 in 200 requests to POST /orders. Logs show 'TypeError: Cannot read properties of undefined (reading "id")' at orders.js:48, which reads req.user.id. The route is behind an auth middleware. The failures started after a deploy that added a 30-second cache on the session lookup. Present your diagnosis and fix as a single HTML page: what you checked, in order, the root cause, the code change, and how you would confirm it is fixed.”
Skill commit: bb92a77

Without the skillWith systematic-debugging

⇆

Without, runWith systematic-debugging, run3 independent runs per side, none selected by hand.

Download the with-skill page

writing-guidelines

Review docs/prose for Writing Guidelines compliance. Use when asked to "review my docs", "check writing style", "audit prose", "review docs voice and tone", or "check this page against the writing handbook".

Category: Testing
Source: vercel-labs/agent-skills
License: Not stated
Proof model: claude-opus-5-5
Brief: “Review and rewrite this getting-started page for Pipewell, a data pipeline CLI. The draft opens with two paragraphs of company history, mixes 'you' and 'the user', buries the install command in the fourth paragraph, uses passive voice throughout, and ends without a next step. Write the improved page as a single HTML document, followed by a short list of what you changed and why.”
Skill commit: 063bee9

Without the skillWith writing-guidelines

⇆

Without, runWith writing-guidelines, run3 independent runs per side, none selected by hand.

Download the with-skill page

We don’t pick a winner for you: the pages are the evidence. Both proofs can be repeated with the brief, model and commit shown above. See alternatives to systematic-debugging or alternatives to writing-guidelines.