Cited, replayable HTTP failure signatures — test any HTTP client against real outages.
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.
Score a battery session and return the report card.
Battery token
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.
Run in a terminal
claude mcp add --transport http faultydata https://faultydata.com/mcp
Score a battery session and return the report card.
| Argument |
|---|
| Type |
|---|
| Description |
|---|
| token* | string | Battery token |
{
"type": "object",
"required": [
"token"
],
"properties": {
"token": {
"type": "string",
"description": "Battery token"
}
}
}Get a specimen's anatomy (the incident story), citation, and replay signature.
| Argument | Type | Description |
|---|---|---|
| id* | string | Specimen ID, e.g. FD-2026-001 |
{
"type": "object",
"required": [
"id"
],
"properties": {
"id": {
"type": "string",
"description": "Specimen ID, e.g. FD-2026-001"
}
}
}List every failure specimen in the FaultyData registry as a compact summary (id, title, category, severity, path, real-world flag, test tags). Use get_specimen for the full story and signature.
No arguments.
{
"type": "object",
"properties": {}
}How to replay a specimen against a client-under-test (URL, curl, params for seed/key).
| Argument | Type | Description |
|---|---|---|
| id* | string | Specimen ID |
{
"type": "object",
"required": [
"id"
],
"properties": {
"id": {
"type": "string",
"description": "Specimen ID"
}
}
}Mint a report-card battery token. The client then runs /run/{token}/{id} per specimen and reports outcomes.
| Argument | Type | Description |
|---|---|---|
| client | string | Client label, e.g. httpx/0.28 |
{
"type": "object",
"properties": {
"client": {
"type": "string",
"description": "Client label, e.g. httpx/0.28"
}
}
}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 20, 2026.
Takes a password, API key or token as an argument, so the secret passes through the model's context and chat logs.
Fix: Read credentials from the server's own configuration (environment, OAuth) instead of a tool argument.
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.