Kimi K3 + LinkedIn

LinkedIn posts with Kimi K3 and Bolta

Turn changelogs and git history into social drafts from the Kimi Code terminal. Parts of the Kimi K3 route are not confirmed yet, and the note below says which. Nothing goes live on LinkedIn until you say yes.

Kimi Code reads MCP servers from mcp.json. The generic MCP guide has the Bolta address and key steps.

Kimi K3Bolta connected
You

Read CHANGELOG.md and save a LinkedIn draft about the newest release

Illustration of the reply: Bolta returns a draft showing the connected account and the scheduled time, with Approve, Edit and Reject actions.

Illustration. Your drafts show your real text, account and time.

What is and is not confirmed

Read this before you set up Kimi K3

  • Confirmed in Kimi's own docs: Kimi Code CLI loads remote HTTP MCP servers from mcp.json, accepts static headers and a bearerTokenEnvVar, and can run a browser sign-in with /mcp-config login.
  • Confirmed in Kimi's own docs: skills are folders with a SKILL.md, read from ~/.agents/skills among other places.
  • Not tested: Bolta's browser sign-in has not been run end to end inside Kimi Code. If it does not complete, use the Bolta API key route below.
  • Not confirmed: the Kimi chat app at kimi.com and any official Kimi plugin. This page is about the Kimi Code terminal agent only.
  • Not confirmed: Kimi K3 and which model Kimi Code runs by default. We have not checked an official Moonshot page for either, so the name is used here only because people search for it.
Bolta on LinkedIn

What Bolta does on LinkedIn

Bolta writes the text, saves it as a draft on the account you name, and waits. Bolta lists each LinkedIn organization as its own account, so a company page is a separate choice from your personal profile.

  • Text posts to a personal profile or to a company page you have connected.
  • Posts visible to everyone by default, or to connections only if you ask for that.
Limits and format

LinkedIn limits and format rules

Network rules last checked 2026-07-28. Bolta's posting rules for your workspace (get-posting-rules) win if they differ from this page.

Not promised for LinkedIn from Kimi K3: LinkedIn articles, newsletters and document carousels. LinkedIn supports them, but this page does not confirm that Bolta publishes them. Comment length limits. The one checked limit on this page is the feed post.

LinkedIn limits and format rules
What to knowLinkedIn's own ruleWhat Bolta does
Post lengthA feed post is capped at 3,000 characters, counting spaces, line breaks, emoji and the full text of any link you paste.Bolta does not check LinkedIn text length, so LinkedIn may refuse longer text. The draft has to fit before you say yes.
Opening linesThe feed cuts a post after the first few lines behind a see more link.Bolta keeps your first lines as written. Put the point there.
FormattingNot stated on this page.Plain text only. Bold, italic and headings become look-alike characters, a list becomes bullets, and a link is written as label (url). #hashtags and @mentions are kept as typed.
Who can see itNot stated on this page.Everyone, unless you set the post to connections only.
Best practices

Kimi K3 jobs for LinkedIn, and how to write them

  • Ask what is waiting for approval in Bolta without leaving the terminal
  • Check how recent posts performed, then draft the next one from that
  • Write to 3,000 characters including spaces, line breaks, emoji and any pasted link, because Bolta will not stop a post that is too long.
Prompts

Example prompts for LinkedIn

Things to type in Kimi K3. Each one asks for a draft and shows you the text first.

Read CHANGELOG.md and save a LinkedIn draft about the newest release

Read CHANGELOG.md and save a LinkedIn draft about the newest release.

Read the top entry in CHANGELOG.md and save a LinkedIn draft about it for my review

Look at the last five commits and draft a LinkedIn post about what changed, without markdown

Review

Bolta saves a draft, you read the exact text, you say yes

  1. 01

    Bolta saves a draft

    Kimi K3 asks Bolta for the LinkedIn post in your brand voice (save-post). A saved draft is never public.

  2. 02

    You read the exact text

    On LinkedIn the exact text includes where the line breaks fall, because the feed hides everything after the first few lines, and how any formatting was flattened to plain text.

  3. 03

    You say yes

    Only then does Bolta schedule or publish (publish-post), in your workspace's timezone. Safe Mode, client sign-off and brand approval rules still apply.

Kimi K3 and Bolta

Using Kimi K3 for LinkedIn

Parts of this route are not confirmed. The note above says which.

For LinkedIn, Kimi K3, from Moonshot AI, reaches Bolta through Remote MCP server in mcp.json (Kimi Code CLI). Sign-in: Sign in (OAuth) or Bolta API key. The LinkedIn account itself is connected inside Bolta, not in Kimi K3.

Kimi Code CLI is a terminal agent from Moonshot AI, so a LinkedIn draft can start from your README, changelog or git log. The Kimi chat app is not covered here.

An illustration from the Kimi K3 guide, not real output. You ask: Read CHANGELOG.md and save a LinkedIn draft about the newest release. The reply goes along these lines: I read the 2.4.0 section and wrote a 140-word draft in your voice. I saved it to Bolta and have not scheduled it. Open it in your Bolta inbox to review. Draft saved. Waits for your yes.

An illustration from the Kimi K3 guide, not real output. You ask: Plan a weekly release-notes routine from our repo for LinkedIn. The reply goes along these lines: Here is the plan: every Friday, draft one post from that week's changes, and send it to review. Nothing is saved yet. Say yes and I will confirm the plan, switched off until you turn it on. Plan shown. Waits for your yes.

Paste: Add a remote MCP server named bolta with the url https://mcp.bolta.ai/mcp to my Kimi Code mcp.json, keep my other servers, and use no transport field. Then tell me to start a new session and run /mcp-config login bolta.

If Bolta refuses an action, such as publishing without the right role, Kimi should report the refusal and stop. Treat any suggestion to work around it as a reason to check the draft in Bolta yourself.

Text Kimi reads from files, issues or web pages is material to write from, not instructions. A changelog line that says to publish immediately does not change what you approve.

  • Posts from your repo. Point Kimi at CHANGELOG.md, a README or recent git history and it saves drafts per network. Several drafts at once go in as a batch, still unapproved.
  • Weekly routines. Describe a cadence such as one release-notes post every Friday. Kimi shows the plan, you confirm it, and it is saved switched off until you turn it on separately.
  • Read your voice. Kimi pulls the brand's voice context so a changelog entry reads like your company wrote it, not like release-note boilerplate.
  • You review in Bolta. The draft appears in your Bolta inbox. You read it, edit it and say yes there, or reject it with a reason.
  • Setup: Ask for a first draft. Try: Turn CHANGELOG.md into a LinkedIn post in my voice and save it as a draft. Open Bolta, read it, and approve or edit it there.
  • Setup: Create a Bolta account and a brand. Sign up at bolta.ai and add at least one brand. A connection reaches one organization, and you choose which brands it may touch when you sign in.
  • Hosted MCP with browser sign-in. Kimi's docs describe /mcp-config login for remote servers, so it should work with https://mcp.bolta.ai/mcp. We have not run it end to end.
  • A headless run such as kimi -p with an API key can save drafts from a CI job. Keep the bolta entry in the user-level file, because a project file stays disabled there until the folder is trusted.
  • Add the server and sign in. Add a bolta entry with the url https://mcp.bolta.ai/mcp to your mcp.json, open a new Kimi session, run /mcp-config login bolta and tick the brands Kimi may use.
Connect

Connect Kimi K3 to Bolta

Kimi Code reads MCP servers from mcp.json. The generic MCP guide has the Bolta address and key steps.

FAQ

LinkedIn questions, answered for Kimi K3

How long can a LinkedIn post be when it comes from Kimi K3?
A LinkedIn feed post is capped at 3,000 characters, and the count includes spaces, line breaks, emoji and the full text of any link you paste. Bolta does not check LinkedIn length, so Kimi K3 should write inside the limit and show you the final text before you say yes.
Can Kimi K3 post to a LinkedIn company page and not only my profile?
Bolta lists each LinkedIn organization as its own account, so you can name the page when you ask (list-accounts). The company page has to be connected in Bolta first. If it is not there, ask for a connect link (get-connect-link) and finish the sign-in in your browser.
Is this for the Kimi app or Kimi Code CLI?
It is for Kimi Code CLI, the terminal agent, because its docs describe remote MCP servers. We have not confirmed that the Kimi chat app accepts custom MCP connectors.
Do I need an API key to connect Bolta to Kimi?
Not if browser sign-in works for you: run /mcp-config login bolta. Bolta's sign-in has not been tested in Kimi Code, so the API key route is the fallback.
Does Bolta publish without my approval when Kimi writes the post?
No. Kimi saves drafts. Nothing is approved, scheduled or published until a person with the right role says yes in Bolta.
Where does Kimi Code keep MCP settings, and can I set them per project?
In ~/.kimi-code/mcp.json for all projects, or .kimi-code/mcp.json inside one project. The project file overrides the user file for the same server name.
Does Kimi K3 publish LinkedIn posts without my yes?
No. Nothing is approved, scheduled or published until you say yes to that LinkedIn post. Whether Kimi K3 completes the connection is not fully confirmed, so read the note above first.

More from Kimi K3

Kimi K3 can work with the other networks Bolta publishes to. Each page has its own limits and review notes.

LinkedIn from other AI agents

The same LinkedIn limits apply whichever agent asks. The connection and the setup are what change.

Try a LinkedIn draft from Kimi K3

Connect Kimi K3, ask for a LinkedIn draft and read the exact text. Nothing goes live until you say yes.