A U.S. voter's ballot by address, with an exact quote and source behind every candidate position.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Requests go through mcptry because browsers can't call most MCP servers directly. Header values are sent to the server for that request only and are never stored or logged.
Tool list from our last check. Connect live to refresh it.
Every researched race on the ballot for a U.S. street address, with each choice's position on the issues that office decides, the exact quote, and the source URL.
Full street address with city, state and ZIP
This tool isn't marked read-only, so it may change something on the server or in an account you connected.
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
claude mcp add --transport http plain-ballot https://plainballot.com/mcp
Every researched race on the ballot for a U.S. street address, with each choice's position on the issues that office decides, the exact quote, and the source URL.
| Argument | Type | Description |
|---|---|---|
| address* | string | Full street address with city, state and ZIP |
{
"type": "object",
"$schema": "https://json-schema.org/draft/2020-12/schema",
"required": [
"address"
],
"properties": {
"address": {
"type": "string",
"maxLength": 200,
"minLength": 5,
"description": "Full street address with city, state and ZIP"
}
}
}The 18 issues Plain Ballot tracks. Each has two sides, each worded the way its own supporters would say it. Use the short side labels ("l" and "r") when giving priorities to score_fit.
No arguments.
{
"type": "object",
"$schema": "https://json-schema.org/draft/2020-12/schema",
"properties": {}
}How closely each choice on the voter's ballot fits the priorities the voter stated, from 0 to 100. Plain arithmetic, the same as plainballot.com: issues where a choice has no sourced position are left out, never guessed. Only use priorities the voter actually gave you.
| Argument | Type | Description |
|---|---|---|
| address* | string | |
| priorities* | array |
{
"type": "object",
"$schema": "https://json-schema.org/draft/2020-12/schema",
"required": [
"address",
"priorities"
],
"properties": {
"address": {
"type": "string",
"maxLength": 200,
"minLength": 5
},
"priorities": {
"type": "array",
"items": {
"type": "object",
"required": [
"issue",
"side"
],
"properties": {
"side": {
"type": "string",
"description": "A side label from list_issues, or \"neither side\""
},
"issue": {
"enum": [
"tax",
"wages",
"trade",
"debt",
"housing",
"schools",
"transit",
"safety",
"sentencing",
"immigration",
"guns",
"abortion",
"gender",
"cannabis",
"health",
"energy",
"voting",
"judicial"
],
"type": "string"
},
"strength": {
"enum": [
"lean",
"strong"
],
"type": "string",
"default": "lean"
},
"importance": {
"enum": [
"low",
"medium",
"high"
],
"type": "string",
"default": "medium"
}
}
},
"maxItems": 18,
"minItems": 1
}
}
}We read every tool's name, description and inputs the way the model will, looking for instructions hidden from you. A clean result means no warning signs, not a guarantee. Scanned Nov 3, 2026.
No readOnlyHint / destructiveHint annotations, so clients can't tell a read from a write without asking.
Fix: Add annotations to each tool, e.g. { readOnlyHint: true } for tools that only read.