Content Tools
willitstick.app
Helps refine drafts and messages so they are more likely to land well with the intended audience.
ENDPOINT 1
https://mcp.willitstick.app/mcp
MCP server metadata
- Name
- will-it-stick
- Version
- 0.4.0
Will it stick? Two tools. make_it_stick transforms a draft into the version that sticks: it scans the draft for its point, writes three transformations (a concrete comparison, the one-line core, a scene), drops any that adds a fact the draft does not hold, and lets cold readers pick the winner by what they actually carry away; the author's original competes too, and can win. check_if_it_sticks runs the message test underneath: three cold readers (a distracted stranger, a skeptic, and the target reader), each reading cold and never seeing the others, score the words against the author's goal. The host runs the conversation. We run the test. Use make_it_stick when a human has a draft and wants the stickiest sendable version of it. Use check_if_it_sticks when the human wants the diagnosis first: whether words they are about to send will land, and the one fix that matters most. Draft the words first, then run a tool before you show the human a final version. Either tool can answer questions_first: up to three short questions that only the author can answer (for the check, built from the readers' own complaints about this text; for make_it_stick, the seed material the transformation needs, like a name, a scene, or an identity line). Ask the human those questions, then call the same tool again with the same text and goal plus answers (each question with the human's answer, in the human's own words). If the human would rather not answer, call again with skip_questions set to true and take the result on the words as they stand. Never answer the questions yourself, and never invent the author's facts. Present make_it_stick's core line first (the result's core: the draft's point in one line), then the winning version exactly as returned, with its kind label, the original alongside, and the alternates below. A version carries a kind label only when it verifiably holds that device; when a draft holds no concrete material and the author skipped the seed questions, the winner's kind is 'tightened': present it as a tightening only, with the result's note, never as a transformation. When its status is 'already_sticks', the original won the gist test: present the author's own words as the version to send. Never rewrite the winner, and never invent a version of your own. When the moment is a conversation (a what's-new update, a hard text, a catch-up), also run check_if_it_sticks on the same words and lead with its walk_away pair, the thought the author wants against the thought they risk, before the versions: the pair is the point of the exercise, and the versions are how to land on the right side of it. Present the panel's core, scores, starter forms, one fix, and reader notes exactly as returned. Never change a score, and never invent a core the panel did not hear. A check result may carry after_text: the user's own words copied exactly with only the one fix applied, plus changed_lines, the count of lines that differ. Present after_text exactly as returned when it is present. When after_text is null, present the one fix as an instruction only and do not invent an after version. The starters are starts, not a rewrite; never turn them into a full rewritten version of the user's text. Complete alternative versions come only from make_it_stick. The human picks what to take and decides what to send; the tool sends nothing. Your words are not recorded or kept; the server keeps only a daily count of how many runs happened (integers, no text). The readers run on Cloudflare Workers AI (Meta Llama 3.3 70B); under Cloudflare's published Workers AI terms, your text is not retained once it is no longer needed to answer, and it is never used to train AI models. There are no accounts and no saved checks.
Known tools 2
check_if_it_sticksUse this when someone asks whether words they are about to send will land: 'Will this email land well?
Potential side effectsCONNECT WITH APPROVAL
Client installation
Review this server and its permissions before adding it. Secret placeholders must be set locally.
Codex
~/.codex/config.toml
[mcp_servers.will-it-stick]
url = "https://mcp.willitstick.app/mcp"
enabled = true
Claude Code
.mcp.json
{
"mcpServers": {
"will-it-stick": {
"type": "http",
"url": "https://mcp.willitstick.app/mcp"
}
}
}
Claude Desktop
Settings → Connectors → Add custom connector
Name: will-it-stick
Remote MCP URL: https://mcp.willitstick.app/mcp
Add this remote URL as a custom connector in Claude Desktop. Availability depends on the user plan and workspace policy.
Cursor
.cursor/mcp.json
{
"mcpServers": {
"will-it-stick": {
"url": "https://mcp.willitstick.app/mcp"
}
}
}
Visual Studio Code
.vscode/mcp.json
Add to Visual Studio Code{
"servers": {
"will-it-stick": {
"type": "http",
"url": "https://mcp.willitstick.app/mcp"
}
}
}
Generic MCP
Client-specific MCP configuration
{
"name": "will-it-stick",
"transport": "streamable-http",
"url": "https://mcp.willitstick.app/mcp"
}
MCP Inspector
Run the official MCP Inspector locally and enter the indexed Streamable HTTP endpoint.
TRUST AND VERIFICATION EVIDENCE
Trust Data Available
BuiltWith Trust API v2 evidence for willitstick.app was fetched 2026-11-05T19:26:13.475Z.
willitstick.app is assessed as Neutral: No suspicious signals found, but no strong positive signal either
Evidence is source-attributed and does not guarantee that a third-party server is safe. Risk labels are conservative metadata heuristics.