Turn what you just built into build in public posts for X, in your voice. Drafts, queue, metrics.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Your Loudcode API key. Create one in Loudcode, Account settings.
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.
Write a build in public post for X about work the user just did, in their own voice (Loudcode knows their product, tone and past posts). Pass a concrete summary of what was built or changed in this session: features, fixes, decisions, numbers, and why it matters to users. Returns the draft (and alternatives). It is saved in Loudcode for the user to review; it is NOT posted.
What was built or changed, and why it matters. Plain language, specific.
single (default): one X post. thread: 3 to 5 posts. long: LinkedIn style.
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 loudcode https://loudcode.dev/mcp \ --header 'Authorization: YOUR_CREDENTIALS'
Headers
Authorization (required): Your Loudcode API key. Create one in Loudcode, Account settings.Write a build in public post for X about work the user just did, in their own voice (Loudcode knows their product, tone and past posts). Pass a concrete summary of what was built or changed in this session: features, fixes, decisions, numbers, and why it matters to users. Returns the draft (and alternatives). It is saved in Loudcode for the user to review; it is NOT posted.
| Argument | Type | Description |
|---|---|---|
| format | string | single (default): one X post. thread: 3 to 5 posts. long: LinkedIn style. |
| summary* | string | What was built or changed, and why it matters. Plain language, specific. |
{
"type": "object",
"required": [
"summary"
],
"properties": {
"format": {
"enum": [
"single",
"thread",
"long"
],
"type": "string",
"description": "single (default): one X post. thread: 3 to 5 posts. long: LinkedIn style."
},
"summary": {
"type": "string",
"description": "What was built or changed, and why it matters. Plain language, specific."
}
}
}List the user's Loudcode drafts waiting for review, and posts queued to go out, with their ids.
No arguments.
{
"type": "object",
"properties": {}
}Post ideas Loudcode found in the user's product, numbers, posts and repo. Good when they shipped nothing new this week.
No arguments.
{
"type": "object",
"properties": {}
}How the user's recent X posts did: views, likes, replies per post and totals, best post first.
| Argument | Type | Description |
|---|---|---|
| days | number | Look back this many days (default 30, max 90). |
{
"type": "object",
"properties": {
"days": {
"type": "number",
"description": "Look back this many days (default 30, max 90)."
}
}
}Queue a Loudcode draft to post to the user's X account: at their next free posting slot, or at a specific time. Only do this when the user asked to queue or schedule it.
| Argument | Type | Description |
|---|---|---|
| at | string | Optional ISO 8601 time with timezone, e.g. 2026-10-14T09:00:00+02:00. Omit for the next free slot. |
| draft_id* | string | The draft's id (from draft_post or list_drafts). |
{
"type": "object",
"required": [
"draft_id"
],
"properties": {
"at": {
"type": "string",
"description": "Optional ISO 8601 time with timezone, e.g. 2026-10-14T09:00:00+02:00. Omit for the next free slot."
},
"draft_id": {
"type": "string",
"description": "The draft's id (from draft_post or list_drafts)."
}
}
}Do the user's weekly check-in from what you know about their week, and Loudcode plans a week of posts (saved as drafts with suggested times, not posted). Answer each question concretely; leave one empty if you don't know.
| Argument | Type | Description |
|---|---|---|
| next | string | What's next, or what are they unsure about? |
| worked_on* | string | What did they work on or ship this week? |
| hard_or_decision | string | What was hard, surprising, or a decision they made? |
| numbers_or_feedback | string | Any numbers or feedback: users, signups, revenue, messages. |
{
"type": "object",
"required": [
"worked_on"
],
"properties": {
"next": {
"type": "string",
"description": "What's next, or what are they unsure about?"
},
"worked_on": {
"type": "string",
"description": "What did they work on or ship this week?"
},
"hard_or_decision": {
"type": "string",
"description": "What was hard, surprising, or a decision they made?"
},
"numbers_or_feedback": {
"type": "string",
"description": "Any numbers or feedback: users, signups, revenue, messages."
}
}
}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 2, 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.