One shared project memory for your AI coding tools: recall, save and brief decisions anywhere.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Personal access token created at https://denasty.me/connect (starts with dn_pat_), sent as a Bearer token.
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.
Deletes one memory by id (from recall). Cannot be undone.
The server marks this tool as destructive. Running it may delete or overwrite data on the server or in any account you connected with your headers.
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http memory https://denasty.me/v1/mcp \ --header 'Authorization: Bearer YOUR_TOKEN'
Headers
Authorization (required): Personal access token created at https://denasty.me/connect (starts with dn_pat_), sent as a Bearer token.Deletes one memory by id (from recall). Cannot be undone.
| Argument | Type | Description |
|---|---|---|
| id* | string |
{
"type": "object",
"required": [
"id"
],
"properties": {
"id": {
"type": "string"
}
}
}Lists the projects DeNasty has memory for, with counts and last update.
No arguments.
{
"type": "object",
"properties": {}
}Returns the user's confirmed personal preferences and writing rules that they allowed to share with AI tools.
No arguments.
{
"type": "object",
"properties": {}
}Returns the project's memory as a Markdown section ready to paste into CLAUDE.md, AGENTS.md or Cursor rules.
| Argument | Type | Description |
|---|---|---|
| project* | string |
{
"type": "object",
"required": [
"project"
],
"properties": {
"project": {
"type": "string"
}
}
}Returns what DeNasty remembers about a project (decisions, conventions, context, open tasks, notes), newest first. Optional text query and kind filter. Omit project to search across all projects.
| Argument | Type | Description |
|---|---|---|
| kind | string | |
| limit | integer | |
| query | string | Words to search for |
| project | string | Project or repository name, e.g. my-app |
{
"type": "object",
"properties": {
"kind": {
"enum": [
"decision",
"convention",
"context",
"todo",
"note"
],
"type": "string"
},
"limit": {
"type": "integer",
"maximum": 100,
"minimum": 1
},
"query": {
"type": "string",
"description": "Words to search for"
},
"project": {
"type": "string",
"description": "Project or repository name, e.g. my-app"
}
}
}Saves one memory for a project so every AI tool of this user sees it later. Use for decisions (and why), conventions, important context, open tasks. Keep it short and self-contained. Secrets are removed automatically.
| Argument | Type | Description |
|---|---|---|
| body* | string | |
| kind* | string | |
| tags | array | |
| title* | string | |
| project* | string | Project or repository name |
{
"type": "object",
"required": [
"project",
"kind",
"title",
"body"
],
"properties": {
"body": {
"type": "string",
"maxLength": 8000
},
"kind": {
"enum": [
"decision",
"convention",
"context",
"todo",
"note"
],
"type": "string"
},
"tags": {
"type": "array",
"items": {
"type": "string"
},
"maxItems": 10
},
"title": {
"type": "string",
"maxLength": 200
},
"project": {
"type": "string",
"description": "Project or repository name"
}
}
}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 22, 2026.