Your assistant writes an introduction from what you have told it and submits it. UpTrust builds your profile from it and publishes your open questions as asks: posts people on UpTrust can answer, labeled “imported” so readers know an assistant drafted them. This publishes right away, with the usual notifications. You can edit or retract each ask afterwards from the page UpTrust links to. Without a connected assistant, you can paste the same introduction at uptrusting.com/start.
Ask it like this
- “Introduce me to UpTrust.”
- “Write my UpTrust introduction from what you know about me.”
When another tool fits better
- Who can help me with this?
You want to look for help without publishing anything.
Before you use it
- It needs the Import your LLM introduction and publish asks for you permission, which UpTrust asks you to approve when you connect.
- It writes to UpTrust, and part of what it does cannot be undone, such as notifications already sent or text it replaces. Its effects reach people or services beyond your own account.
- In Claude, connect
https://mcp.uptrusting.com/mcp. In ChatGPT, connecthttps://mcp.uptrusting.com/chatgpt. How to connect.
What your assistant is told
Your assistant decides when to use this tool from the title and description UpTrust’s server gives it. Here they are exactly as the server sends them.
- Tool name
submit_llm_introduction- Title
- Introduce me to UpTrust (this publishes posts)
- Description
Use this when the user says "introduce me to UpTrust", "set up my UpTrust profile", or "join UpTrust". Write their introduction from what they share in this conversation, not earlier chats or saved memory. If that is too thin, ask them for a few lines on what they care about and what they are trying to figure out, then write it. Include only what you are confident about, and leave out anything they would consider private. It writes their UpTrust profile from that introduction and publishes their open questions as asks: posts people on UpTrust can answer. To look for help without publishing anything, use find_help instead. Submit the raw text of the introduction you wrote to the user's own network profile (or the output of the introduce_me_to_uptrust prompt). Pass the text exactly as written; the server parses the sections itself. THIS IS A LOUD, IMMEDIATE WRITE, not a draft or a preview: submitting PUBLISHES up to 5 of the introduction's open questions as ask posts right away, with the normal notifications those posts send. Each is labeled "imported" so readers know an assistant drafted it, and the user can edit or retract any of them from the payoff page at /import. Call it once per introduction. You do NOT need to read the introduction back to the user or collect their approval of the text first — write the best introduction you can from what they share in this conversation and submit it. They review and edit on the payoff page afterwards, and can retract any ask with one tap. The response is the payoff, structured: every parsed ask with its status and post URL, what the graph can ALREADY serve for each one (read `strength` — a "weak" match is a suggestion, never an answer), the topics the introduction places them near, and a `speakable` summary. `autoPublished: false` means nothing was published this time — because a new account spent its daily cap of auto-published asks, or the introduction parsed no publishable asks — a normal outcome, not a failure; check each ask `status`. The equivalent manual path is the paste-back web flow at https://uptrusting.com/start?src=llm, which the user can always use instead. The introduction is the whole job. Afterwards, and only optionally, `followUp` invites you to call suggest_profile_updates; the introduction never depends on it.