Loading...
Grok Build/TarenoGrok Build is xAI's terminal coding agent, and the source is published in xAI's own repository.
Grok BuildGrok Build / Tareno
Watch what happens right after you connect Tareno: your assistant can see your social accounts, and you can start writing.
Grok BuildIllustrated demo · no live account
Run the grok mcp add command with the http transport, a name such as tareno and Tareno's MCP address, or add the server under mcp_servers in ~/.grok/config.toml. Use the project scope only for servers without secrets. For OAuth, Grok opens a browser flow and stores the token. xAI documents the flow, and Tareno has not completed a client test.
xAI's guide says grok mcp add writes to your user file at ~/.grok/config.toml by default, and writes to .grok/config.toml in the repository when you pass the project scope. A server without secrets can live in the project file and be committed; a personal sign-in belongs in the user file.
Pass the http transport, a name made of letters, numbers, hyphens or underscores, and the address shown on the right. xAI recommends this native URL form over wrapping a server in an npx mcp-remote proxy, because Grok then handles OAuth itself.
The guide says Grok opens a browser authorization flow when a server asks for OAuth, and you can also press i on the server in the /mcps screen. Tokens are stored as plaintext in ~/.grok/mcp_credentials.json with owner-only permissions, so keep that machine yours.
grok inspect shows which servers are loaded and where they come from, and grok mcp doctor is the first diagnostic step. Then ask Grok Build to call your connected accounts. Your account appearing in the answer is the confirmation; an empty answer means it has to be connected inside Tareno.
"url": "https://tareno.co/api/mcp"Input
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 sits next to the work that people want to talk about. These jobs use that proximity, and each has a workflow guide.
Grok Build can read the commits and the changelog of a tagged release and write the post from what changed, not from memory. You read it against the diff, then it is saved as a draft, which is more reliable than writing the announcement an hour after the build when details have blurred.
When a migration or an optimization is done, the evidence is still in the repository: before and after numbers, the issue thread, the benchmark script. Grok Build can pull those into a LinkedIn draft with the limits stated, leaving you to remove anything private before it is saved.
xAI's guide says Grok also reads MCP definitions from ~/.claude.json, .cursor/mcp.json and a project .mcp.json, with config.toml taking priority. A team that already configured Tareno for another coding agent may find it picked up, which is convenient and worth checking with grok inspect.
One guide for every network this agent can work with in Tareno, each with its own example, prompts and limits.
Codex, Claude Code and Cursor reach Tareno through documented setups that Tareno knows well. Grok Build uses the same remote MCP idea with its own commands and config file, plus compatibility with other tools' definitions. If you already work with xAI's models, it keeps the announcement in the same session as the code. If not, choose the agent your team already uses.
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:
No. Grok Build is xAI's terminal coding agent, while Grok Bot is a separate product with a plugin and connector model of its own. Tareno has a separate guide for each, and the setup steps on one page do not apply to the other.
In ~/.grok/config.toml for your user, and in .grok/config.toml inside a repository for project servers, according to xAI's guide. Project files are limited to MCP servers, plugins and permission rules, and secrets should not be written into a committed one.
Yes, according to xAI's guide: when a server requests OAuth, Grok opens a browser authorization flow and stores the tokens. You authenticate once, and the stored credentials are reused, so sign in with your own Tareno account rather than a shared one.
Not yet. The setup follows xAI's published documentation and Tareno's OAuth is standard, but a full read-only run through Grok Build has not been completed. Run `list_accounts` first and treat the result as the test for your own setup.
xAI's guide names it as the first diagnostic step when a server does not load. Together with grok inspect, which shows loaded servers and where each one came from, it tells you whether a problem is in the config, the transport or the sign-in.
No. A scheduling or publishing request becomes an approval action in Tareno, and nothing reaches X or LinkedIn until you approve it in the dashboard. Grok Build can only report that the action is pending and read its status afterwards.
Possibly. xAI's guide says Grok also reads ~/.claude.json, .cursor/mcp.json and a project .mcp.json, with its own config taking priority. Run grok inspect to see what it actually loaded, because a stale definition can point at an old address.
Usually `list_accounts` to confirm the destination, then `create_drafts` to save the text. `schedule_posts` comes last and returns an approval action instead of publishing anything, and `get_action_status` tells you what became of the request.
Start with your Tareno account, connect your agent and prepare drafts. You approve what goes live.