ShowTrust for AI-agent users
If you build with Claude Code, Cursor, or any AI coding agent, you don't write integration boilerplate any more — you describe what you want and the agent does it. ShowTrust is built the same way: install the skill, tell your agent 'add a wall of love to my landing', and it ships.
The Problem
Most SDKs assume you read the docs end-to-end — your agent shouldn't have to
Boilerplate-heavy installs blow your agent's context window on nothing
Auth flows, env vars, and config files mean three tabs and a manual checklist
Marketing widgets are an afterthought — you ship features first, proof later
The Solution
A first-party Skill your agent loads in one command
Docs written for agents — JSON-LD examples, env contracts, and snippets
The agent signs you up from your email, configures the project, edits the right file
Same product, same widget — just installed in 30 seconds instead of an afternoon
From inside your editor, you tell the agent: 'install ShowTrust and add a wall of love to my landing page hero, minimal style'. The agent loads our skill, hits the install endpoint with your email, picks up the project slug, edits your landing component, and runs the build.
You verify it works in your dev server. The whole loop — from prompt to live widget — is under 30 seconds. The same skill knows how to add the submission link to your customer success emails, and embed walls on different routes.
Any agent that can run shell commands, edit files, and read documentation. We've tested with Claude Code, Cursor, Cline, and Continue. The skill is just a markdown contract — agents that follow it will work.
A bundled set of instructions, examples, and tool descriptions that an agent loads on demand. It tells the agent exactly how to install, configure, and embed ShowTrust without you having to spell each step out. See [docs/skill](/docs/skill) for the full spec.
By design, narrow. The skill changes only what you ask it to: a single React/Astro/HTML file for the embed and one env var for the project key. It never touches credentials, auth, or unrelated files. Every action is visible in your agent's diff before commit.
We have both. The hosted dashboard works fine in a browser. But if you already live in your terminal and trust your agent for production work, why open three tabs to install a marketing widget? Same product, less friction.
Treat it like any other agent edit — review the diff, revert if it's off. The skill is deliberately small-surface: bad outputs are visible and easy to undo. Nothing happens server-side that you can't roll back from the dashboard.
Hosted submission page + embeddable wall of love. Free while in early access — no card to start.
Ship social proof before your launch
Hosted submission form + one-line embed. Beta testers become a wall of love in time for your Product Hunt day. Free in early access.
Lift conversion on landing & pricing
Authentic social proof on the pages where prospects decide. Embed a wall of love and raise conversion without raising spend.