A curated registry of agent stacks: search by job, get the tools, prompt, and install plan.
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.
How to install a stack: the MCP servers to configure, the shell commands to run, and the notes (a path, a harness command) for the rest. Deciding and configuring are open; the agent's host runs the commands.
The stack's slug.
The stack's namespace owner (a login, or runchr).
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 runchr https://api.runchr.com/mcp
How to install a stack: the MCP servers to configure, the shell commands to run, and the notes (a path, a harness command) for the rest. Deciding and configuring are open; the agent's host runs the commands.
| Argument | Type | Description |
|---|---|---|
| slug* | string | The stack's slug. |
| owner* | string | The stack's namespace owner (a login, or runchr). |
{
"type": "object",
"required": [
"owner",
"slug"
],
"properties": {
"slug": {
"type": "string",
"description": "The stack's slug."
},
"owner": {
"type": "string",
"description": "The stack's namespace owner (a login, or runchr)."
}
}
}The manifest a loader runs: the tools, the model, the harness, and the prompt. This is the thing to install.
| Argument | Type | Description |
|---|---|---|
| slug* | string | The stack's slug. |
| owner* | string | The stack's namespace owner (a login, or runchr). |
{
"type": "object",
"required": [
"owner",
"slug"
],
"properties": {
"slug": {
"type": "string",
"description": "The stack's slug."
},
"owner": {
"type": "string",
"description": "The stack's namespace owner (a login, or runchr)."
}
}
}One stack's card: its copy, its tools, its tier, and its revision.
| Argument | Type | Description |
|---|---|---|
| slug* | string | The stack's slug. |
| owner* | string | The stack's namespace owner (a login, or runchr). |
{
"type": "object",
"required": [
"owner",
"slug"
],
"properties": {
"slug": {
"type": "string",
"description": "The stack's slug."
},
"owner": {
"type": "string",
"description": "The stack's namespace owner (a login, or runchr)."
}
}
}Find agent stacks by problem domain or text. Returns the newest first. Each stack is a runnable combination of registry tools.
| Argument | Type | Description |
|---|---|---|
| limit | integer | How many stacks to return (default 10). |
| query | string | Text matched against the stack name and summary. |
| topic | string | A problem domain id, e.g. software-devops. |
{
"type": "object",
"properties": {
"limit": {
"type": "integer",
"description": "How many stacks to return (default 10)."
},
"query": {
"type": "string",
"description": "Text matched against the stack name and summary."
},
"topic": {
"type": "string",
"description": "A problem domain id, e.g. software-devops."
}
}
}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 6, 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.