Loading...
Claude Code/TarenoClaude Code already sits next to your repository, so it knows what shipped.
Claude Code / Tareno
Watch what happens right after you connect Tareno: your assistant can see your social accounts, and you can start writing.
Claude CodeIllustrated demo · no live account
Use the published Tareno plugin from Claude's plugin directory and sign in with your Tareno account, or add Tareno yourself as a remote MCP server with an API key. Then ask Claude Code to call your connected accounts. If your social accounts appear, the connection works and you can start with a draft.
The plugin route signs in with OAuth and bundles the Tareno social-operations skill. The manual route adds a remote MCP server by hand and authenticates with an API key from your Tareno settings. Choose the plugin unless you need a scripted, repeatable setup on several machines.
For the plugin, search for Tareno in Claude's plugin directory, install it and complete the Tareno sign-in. For the manual route, copy the Claude Code command from the MCP docs page and keep the key in the TARENO_API_KEY environment variable instead of pasting it into a prompt.
Run the check on the right to confirm that Claude Code knows the server, then start a fresh session so the tool list is reloaded. A connection that exists but shows no tools usually needs that restart.
Ask Claude Code to run your workspaces and your connected accounts and to read each network's rules for the account it will use. The bundled skill does this on its own, but saying it once makes the first draft predictable.
claude mcp get tarenoInput
Provide the approved source and audience.
Review-ready drafts
Use Tareno and compare the returned draft IDs.
Approval in Tareno
Review text, media, account, and time in Tareno and approve.
A terminal agent with repository access is at its best when the source of a post is a file or a Git range. These four jobs follow that pattern, and each has its own workflow page.
Point Claude Code at the section of CHANGELOG.md for the new version and the commit range behind it. It separates fixes, features and breaking changes, keeps each post short enough for readers who do not know the project, and saves every post through saving drafts as its own draft. Tareno does not chain posts into an X thread through MCP, so you link the replies yourself in X or publish them as independent posts.
When a migration or a performance fix taught you something, give Claude Code the notes and the evidence you are allowed to share. It writes the problem, the decision and the lesson in a professional register, and flags every number that has no source in your notes so you can remove it before approval.
Open-source updates work when the people behind a release are named. Claude Code can read the contributor list from your history, draft the update with links to the release and the issues, and ask you to confirm each name before saving the draft.
Once the drafts are final, Claude Code can request a time with requesting a time. Tareno answers with an approval action and a link, so the request sits in the dashboard until you approve it, and the approval status tells you afterwards what happened.
One guide for every network this agent can work with in Tareno, each with its own example, prompts and limits.
Choose Claude Code when the source is a repository: releases, commits, design notes and pull requests. Choose Claude Cowork for document-heavy jobs such as turning a campaign folder into reviewed deliverables. Choose the Claude app for conversations and one-off drafts. All three use the same published plugin and the same Tareno account, so a draft started in one place can be reviewed in the dashboard from any other.
https://tareno.co/api/mcpUse the same Tareno account across supported clients. Open the relevant setup guide for authentication, first use, reconnect, and revocation.
list_accountsChecked:
Yes. Give it the version section and the commit range, say who the readers are, and it drafts one or several separate posts and saves them through Tareno. It can group fixes and features, but you remain responsible for checking that every claim matches what actually shipped.
Not with the plugin, which signs in through OAuth with your Tareno account. The manual remote MCP route uses an API key from your Tareno settings. Keep that key in an environment variable and never paste it into a prompt or a repository file.
Yes. The Tareno plugin is published in Claude's plugin directory and works in Claude Code, Claude Cowork and the Claude apps. After connecting your own Tareno account, run list_accounts and check the returned profiles before creating a draft or requesting a publishing time.
No. Scheduling and publishing requests return an approval action and a link into Tareno, and nothing reaches a social network until you approve it there. Claude Code should report that the action is pending rather than claim it is done.
Any setup that runs Claude Code without a person present could prepare drafts, but Tareno still requires a human approval before publishing. We have not documented a CI recipe, so test with read-only prompts first and keep credentials in your pipeline's secret store.
The common sequence is `list_accounts` to confirm the destination, `get_platform_schema` for the rules of that account, `create_drafts` to save the text and `schedule_posts` for the request. `get_action_status` closes the loop once you have approved the action.
Tell Claude Code which sources are public, ask it to list every name, number and customer reference it used, and read the draft in Tareno before approving. Because Tareno only sees what Claude Code sends, the check is a quick read of the draft itself.
Claude Code works from a terminal with your repository, which suits releases and technical posts. Claude Cowork is built for longer jobs that work through documents and campaign materials. Both use the same plugin and the same account, so you can pick the one that matches where the source material lives.
Start with your Tareno account, connect your agent and prepare drafts. You approve what goes live.