- Home
- Skills
- Testing & Debugging
- PWA Dev Agent — Mobile-First Build & Release
PWA Dev Agent — Mobile-First Build & Release
Help your coding agent plan, fix, test and safely ship PWAs. Includes an offline Python audit, iPhone/Safari checklists, release gates, review templates and a real-world case study.
$12
Works with the AI Agents you already use
PWA Dev Agent — Mobile-First Build & Release
Example session with this skill installed
In BiblioPromptos, remove the label "Prompt gerado automaticamente" above the generated prompt. Update both app.js and the single-file sitecase.html version. Preserve the Copy action, other buttons and behavior. Run local tests and show me the visual result before publishing. Do not deploy until I explicitly approve.
- Read your context and instructions
- Compiled the pwa dev agent
- Generated the app files
Real case study — BiblioPromptos (4 Nov 2026)
Request: Remove the visible label "Prompt gerado automaticamente" from the generated-prompt card in two versions of an existing PWA-oriented web app. Keep the Copy button at the right, preserve all other behavior, and show the result before publication.
Work performed: An AI coding agent read the PWA Dev Agent skill, inspected the repository and modified only app.js and sitecase.html in an isolated checkout. It removed the heading and right-aligned the existing Copy action.
| Check | Actual outcome |
|---|---|
| JavaScript syntax | PASS — node --check app.js |
| Inline HTML script syntax | PASS |
| Removed phrase | 0 occurrences in both files |
| Copy action | Present in both files; manually tested in Firefox |
| Visual review | Result displayed to the user before publication |
| iPhone Safari | NOT TESTED — no claim of iPhone validation |
| Production deploy | Not performed until the user explicitly approved it |
After approval: The specific change was committed and pushed to GitHub, then manually deployed to Cloudflare Pages. The production JavaScript was checked with SHA-256 and HTTP 200. GitHub evidence: https://github.com/viniciusnevesdev/bibliopromptos/commit/f22376c2d76281db7d9a9f63964d8ad55770fd4b
This is a documented workflow example, not a guarantee that every coding agent automatically has GitHub, browser, or hosting access. The skill supplies instructions, checklists and an offline audit; the buyer provides the actual tools and permissions.
sitecase.html
ZIP · project files
Example file from a real run - the skill writes it into your workspace.
Connects securely to your tools. The creator never sees your data.
About this skill
Build and maintain PWAs without turning every small change into a risky deploy.
PWA Dev Agent is an approval-first workflow for AI coding agents working on Progressive Web Apps. It guides your agent through inspecting a codebase, planning a focused change, working in isolation, testing actual behavior, and presenting a clear review before any production publication.
Includes an offline Python 3 PWA source auditor (no extra dependencies or network), six tests, mobile Safari and iPhone checklists, security and deployment guidance, review templates, and a real Codex-assisted case study with a publicly verifiable GitHub commit.
Designed for mobile-first interfaces, service worker caching, manifests, offline behavior, touch and responsive layout concerns. Supports remote development workflows when your own agent platform provides remote jobs.
Important: This is a skill/instruction package, NOT a hosted coding agent, API key, remote server, browser automation service, or guarantee of Safari installation. Actual repo edits, browser checks and deployment require tools and permissions in your own environment. Publishing requires explicit human approval.
Real-world example: the agent removed a redundant heading from BiblioPromptos in two source distributions, preserved the Copy action, ran checks and previewed the result. Only after approval was the change published to GitHub and verified on Cloudflare Pages.
Case study commit: https://github.com/viniciusnevesdev/bibliopromptos/commit/f22376c2d76281db7d9a9f63964d8ad55770fd4b
Heuristic audits do not replace browser certificates. Visual checks require agent browser access. Task persistence depends on your specific remote execution runtime.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 13 days ago
Requires