flow-redesign
Skill1 starsChanged 47 days agoMITView on GitHub
Measured upgrade of an existing interface. Use to modernize or polish shipped UI that looks dated or inconsistent without breaking URLs, content or behavior. Hands off to `flow-design`, `flow-ui` or `flow-a11y`.
In this repository38 other files it gives its agents.
- AGENTS.mdAGENTS.md
- llms.txtllms.txt
- flow-a11ySkill
- flow-adversarial-reviewSkill
- flow-animateSkill
- flow-apiSkill
- flow-benchmarkSkill
- flow-brainstormSkill
- flow-commitSkill
- flow-data-migrateSkill
- flow-depsSkill
- flow-designSkill
- flow-diagnoseSkill
- flow-docsSkill
- flow-git-fixSkill
- flow-handoffSkill
- flow-migrateSkill
- flow-observeSkill
- flow-onboardSkill
- flow-orchestrateSkill
- flow-planSkill
- flow-pr-fixSkill
- flow-prototypeSkill
- flow-prune-agent-setupSkill
- flow-refactorSkill
- flow-releaseSkill
- flow-repo-briefSkill
- flow-reviewSkill
- flow-shipSkill
- flow-simplifySkill
- flow-specializeSkill
- flow-stylesSkill
- flow-testSkill
- flow-triageSkill
- flow-uiSkill
- flow-verifySkill
- flow-workSkill
- flow-write-skillSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 47
After each step, re-render the baseline screenshots, run the repository checks, and confirm the preserved contracts still hold.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Redesign
Improve what exists without breaking what works. Measure first, change in small reversible steps, and prove each step made things better.
Phase 1: Classify and baseline
- Mode: polish (keep the system, fix execution), refresh (evolve tokens and components),
or overhaul (new direction through
flow-design). Default to the least disruptive mode that meets the goal, and confirm an overhaul with the user. - Baseline: screenshots of key pages at 375, 768 and 1280 in each theme, key states, and any available metrics (Core Web Vitals, accessibility scan results, analytics goals).
- Contracts to preserve: URLs and routes, navigation labels, form field names and order, analytics and test IDs, SEO titles and metadata, logo and legal copy. Change one only when asked, and list it.
- Stack: framework, styling approach, component library and tokens. Work inside them; never migrate frameworks or styling systems as part of a redesign.
Phase 2: Audit
Score each dimension 1 to 5 with evidence (screenshot, selector, measured value) using
references/audit-rubric.md: hierarchy, typography, color and
contrast, layout and spacing, components and consistency, states, responsiveness,
accessibility, performance, and content. Check the generic tells from flow-design.
Turn findings into a punch list: issue, evidence, impact, effort, risk. Order by impact over effort, and keep risky or cross-cutting changes for last.
Phase 3: Upgrade in steps
Work in this order, because each step makes the next one cheaper and less risky:
- Tokens: type scale, color roles, spacing and radius, fixed at the source so every screen improves.
- Typography and spacing: hierarchy, line length, vertical rhythm, alignment.
- Color and surface: one accent, consistent neutrals, contrast in every theme, elevation only where meaningful.
- Components and states: shared components first; add missing hover, focus, loading, empty and error states.
- Layout and composition: replace generic sections only where the content deserves a better structure.
- Motion and detail last, and only with a purpose.
After each step, re-render the baseline screenshots, run the repository checks, and confirm the preserved contracts still hold. Keep steps small enough to review and revert on their own.
Phase 4: Report
Show before and after screenshots per step, the rubric re-scored, contracts confirmed unchanged, and the punch-list items deferred with reasons.
Anti-patterns
- Rewriting from scratch when targeted fixes would do.
- Restyling one page while the shared components and tokens stay broken.
- Swapping fonts, icon sets or libraries without checking licensing, loading cost and existing usage.
- Silently renaming routes, labels or form fields.
- Declaring it better without a before-and-after comparison.
Done when
- The punch list's in-scope items are done, each with before-and-after evidence.
- Rubric scores improved on the targeted dimensions and none regressed.
- Preserved contracts are verified, and deferred items are listed with reasons.
Set a new direction with flow-design, build new surfaces with flow-ui, and finish with flow-a11y.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

