Marketplace
VVYREN code
MCP servers
Tells
Freeno reviews
Install
mcp.json — Claude Desktop, Cursor, Windsurf, VS Code
{
"mcpServers": {
"tells": {
"type": "http",
"url": "https://devconnectplatform.com/api/tells"
}
}
}Claude Code
claude mcp add --transport http tells https://devconnectplatform.com/api/tellsAsk an AI to review its own work and it says the work is good. That is not laziness. The output sits exactly where that model considers normal, so the fingerprint is invisible from the inside.
Tells is the instrument from outside. It counts what gives an AI-built product away.
COPY. Dash density, marker words, and the sentence rhythm that drafted text cannot hide. Human writing is uneven. Nine words, then thirty, then a fragment. Generated writing is smooth. Smoothness is the one signal that resists faking, because raising it means rewriting rather than swapping words.
DESIGN. How many distinct spacing values, type sizes, colors, radii and shadows your interface actually uses. A design is a vocabulary used consistently. A generated one is an arbitrary pile of values that happen to look fine, and the difference between eight spacing values and forty-seven is not a matter of taste. Plus line length, alignment, real WCAG contrast, and whether anything on the page stands out at all.
MOBILE. Tap targets under 44px. Buttons that overlap on a 360px screen. Elements past the viewport edge, and text cut off without an ellipsis.
VISIBILITY. Whether answer engines can read, quote and attribute your pages, or whether one wildcard in robots.txt has been quietly blocking them the whole time.
Every finding carries a number and the baseline it was measured against. Nothing is a model judgement, so the same input returns the same answer twice. Mechanical fixes come back as patches. The rest is flagged and left alone, because rewriting a sentence needs somebody who knows what the product does.
It also states the score it expects after the fix, then checks whether it was right. When the prediction fails, that is reported as a fault in this tool and not as a finding about your work.
STATUS: in development. The signal library and its tests are written. Baselines are being measured against a real corpus before anything ships, because a guessed baseline is not a baseline. The endpoint appears on this page when it answers.
Free. No account, no API key.