with one click
lazyweb-skill
lazyweb-skill contains 9 collected skills from aboul3ata, with repository-level occupation coverage and site-owned skill detail pages.
Skills in this repository
Lazyweb is the design-evidence skill for AI coding agents. Use it before designing, critiquing, or changing product UI when the agent needs real app screenshots, competitor references, best practices, quick examples, creative cross-category ideas, paywall optimization guidance, or mobile growth and monetization A/B test context. It can also route explicit requests to update the local Lazyweb skill pack. It routes to the right Lazyweb mode and tells the agent to use Lazyweb MCP tools instead of guessing from generic training data.
EXPLAIN how something works with a hosted, tappable flow diagram — a walkthrough, a failure trace, an answer to "how does X work?", a hypothetical — WITHOUT touching the product's canonical flow chart. Saves via lazyweb_explain_flow to separate insert-only storage, hosted at /explainer/<id>/. Use THIS (never the generate/save-flowchart path) whenever the goal is explanation rather than recording the product's current state. Trigger on: "explain this with a diagram", "walk me through how X works", "diagram why this failed", "explain using the flowchart", "show me how the pieces fit", "trace what happened".
Generate a canonical FLOW CHART of how a product's agent/app and its backend talk, in the Lazyweb metrics-tab style (per-capability sections, optional Initialization, sync one-round-trip vs async — poll / stream / webhook / queue — on an actor-colored spine), and SAVE it as the product's static flow chart. Run this FIRST — before any proposal — so proposals annotate a stable, shared chart. Trigger on: "generate a flow chart", "map how X talks to its backend", "create the architecture/flow diagram", "flow chart for my product", "diagram the agent<->backend flow", "make the canonical flow chart".
UPDATE a product's saved Lazyweb flow chart to match the current code: look at the existing chart, pull the code diffs since it was last updated, and upsert the chart in place (stable id + URL) with the real changes. A DIRECT update — no proposal / Accept-Decline overlay. This is what the "Update this chart" banner on a /flowchart/<id>/ page copies to your clipboard. Trigger on: "update the flow chart", "refresh the flowchart", "the chart is stale", "sync the flow chart with the code", "bring the diagram up to date".
Update the local Lazyweb skill pack from GitHub and reinstall Lazyweb skills into supported local coding clients and agentic IDEs. Use when the user asks to update Lazyweb, refresh stale Lazyweb slash commands, sync local skills, reinstall the skill pack, or make Codex/Claude/Cursor/OpenCode/Kiro/Factory/ Slate/Hermes pick up the latest Lazyweb skills.
Propose UI/flow changes (or explain context) on a STRUCTURED DIAGRAM that the user reviews and Accepts/Declines on a hosted lazyweb.com page, then apply the accepted ones. Use when the user wants a reviewable set of proposed changes to a product's UI, architecture, or flow — anything where you'd otherwise list suggestions in chat and want the user to approve/reject each before you act. Trigger on: "propose changes", "suggest UI changes for review", "let me review your suggestions", "annotate this flow/diagram", "create a proposal", "review and approve these changes", "explain this architecture/flow visually".
Optimize, improve, or design any product screen — routes on objective. The user-facing umbrella for product-screen work, mobile or web (paywall, pricing, landing, signup, onboarding, dashboard, settings, …). Pick the `objective` INTENT-FIRST, not by whether an image exists: `optimize` (move a conversion metric on an EXISTING screen) and `improve` (raise design quality of an EXISTING screen) both require a screenshot — captured on the user's behalf; `create` (a NEW screen from scratch) routes to the internal lazyweb-design-create backend. For optimize, the server runs the internal pipeline — labels the screen, retrieves structural experiment twins, diagnoses frictions, synthesizes a slot-diverse portfolio of falsifiable hypotheses bound to real A/B experiments — then renders the dark report; the skill captures the screen, generates one mockup per winner, and hosts it. Use when the user wants to redesign, improve, critique, optimize, or design a product screen for conversion or quality.
Use the Lazyweb search MCP tool directly for quick industry references before designing or changing UI. Does not generate a report. Use when the agent needs a lightweight best-practice check from real app screenshots, wants to inspect search coverage, or should grab a few references before building. Trigger on: "quick search", "search Lazyweb", "use lazyweb_search", "quick industry references", "check examples before designing", "look up UI references first".
INTERNAL create / greenfield backend. NOT a user-facing slash command — it is reached via `/lazyweb-design` with `objective=create`, which redirects here for designing a NEW product screen from scratch (no existing screen to ground on). Combines Lazyweb's screenshot database with web research and produces a prototype-first HTML report with side-by-side prototypes and a clustered inspo map. Do not route users here directly; route them to `/lazyweb-design` and let the create objective hand off to this backend.