Skip to content

getting-started.md

Latest commit

 

History

History
97 lines (75 loc) · 4.06 KB

getting-started.md

File metadata and controls

97 lines (75 loc) · 4.06 KB
Raw

Getting Started

Part of the TeamAI CLI Usage Guide.

What you are doing Section
Creating the team repo Set up a team
Joining a repo that already exists Join a team
Share a skill, rule, or other AI asset you wrote with the team Share with the team

Set up a team

  1. Create an empty repository on GitHub, GitLab, GitCode, CNB, TGit, or a private Git host. <team-name>-teamai is a clear name. Give teammates permission to push branches, and leave the default branch protected. No repo yet? Fork one from teamai-hub.

  2. Initialize in the project where you use your AI tool.

    In your AI tool:

    Install the teamai skill: https://github.com/Tencent/teamai-cli/tree/main/skills/teamai , load the teamai skill, then set up TeamAI for my team from scratch.

    Or in a terminal:

    cd /path/to/my-workspace
teamai init https://github.com/your-org/your-repo

    init detects the Git host, signs you in if needed, registers you as a member, installs a session-start hook in the AI tools it finds, and pulls.

  3. Check it.

    teamai doctor      # every line should pass
teamai status      # empty right after init

  4. Publish one file so the first pull is not empty. Put a skill in ~/.claude/skills/<name>/SKILL.md (or a rule in ~/.claude/rules/<name>.md) and run teamai push. Merge the pull request it opens.

  5. Send the repo URL to the team.

More management features are in Distribution mechanism.

Join a team

You need the repo URL.

  1. Initialize in the project where you use your AI tool.

    In your AI tool:

    /teamai Help me join my team's TeamAI, repo URL is https://github.com/your-org/your-repo

    Or in a terminal:

    cd /path/to/my-workspace
teamai init https://github.com/your-org/your-repo

    Add --scope user to install into your home directory. Every project on this machine then sees the files. Without that flag, files install into this project.

    Inside a Git repository, init uses the repository root even when run from a subdirectory. If that root already has a workspace config, run init at the root to change it; a plain subfolder cannot hold a separate config, even with --force. A nested repository or linked worktree has its own workspace root.

  2. Check it.

    teamai doctor
teamai list        # the team's skills, rules, docs, env, agents, hooks and MCP servers

  3. Open your AI tool in that project.

Day-to-day commands are in the Member Guide. Searching what teammates learned is in Team Context.

Share with the team

You have a skill, rule, agent, or MCP server that teammates should have.

  1. Push it.

    In your AI tool:

    /teamai Share my <name> skill with the team

    Or in a terminal:

    teamai push                       # pick from what it finds under your AI tools
teamai push --skill ~/.claude/skills/<name>

    push fills in missing SKILL.md frontmatter, pushes a branch, and opens a pull request on the team repo.

  2. A reviewer on the team repo merges the pull request.

  3. The next session on a teammate's machine pulls it. teamai list skills --source repo shows it before that session.

The pull request is merged, and teamai list skills --source repo shows the skill.

Roles and namespaces are in Distribution mechanism. The format of each asset is in Core Harness.

If a command fails

Run teamai doctor. It names a missing hook, a tool it did not detect, or a token that is not set, and prints a fix. Other cases are in Uninstall & FAQ.

Where to go next

Next Read
Talk with /teamai teamai skill
More management features Distribution mechanism
How to write skills, rules, and the rest Core Harness
After you have joined Member Guide