How many Rovo credits you use depends on the value of each operation. Connecting sources, metadata, tool discovery, and writes are always Free; basic single-product reads are Standard and included with your plan; the vast majority of Enriched TWG API calls are 1-10 Rovo credits per call, so richer answers cost a little more than simple lookups. Your plan includes a monthly Rovo credit allowance, with any extra usage billed transparently.

Enriched commands require Rovo Credits because they aggregate responses from multiple Atlassian and non-Atlassian apps and may use AI inferencing to generate insights.

For example, twg collaborators provides insights on collaborators by analyzing common artifacts and interactions between users in an organization. Similarly, twg context aggregates information across multiple Atlassian and non-Atlassian systems to deliver context around a work item, Confluence page, or user.

To see which commands require Rovo Credits, refer to the Teamwork Graph CLI Command Catalog. To check an individual command in the CLI, run twg help <command>; Enriched commands are labeled [Enriched] in the help output.

For thresholds, plan-level allotments, and what happens when you approach a limit, see Rovo usage limits.

Rovo credit consumption from Teamwork Graph CLI is included in your Rovo credits dashboard in the Atlassian Admin hub. To view it, go to Atlassian Administration, then select Insights > Platform Usage > Rovo Credits.

How is Teamwork Graph CLI different from Atlassian CLI?

Atlassian CLI was built for human-driven administration before agentic usage was common. Teamwork Graph CLI is purpose-built as an agent-first CLI that standardizes commands and outputs for how agents actually work.

Teamwork Graph CLI also offers capabilities beyond Atlassian CLI, including access to cross-product data from the Teamwork Graph and additional Atlassian tools such as Bitbucket, Teams, Customer Service Management, Jira Service Management, and Talent.

Does Teamwork Graph CLI work offline?

No. Teamwork Graph CLI is a local binary, but it accesses cloud data and services. A network connection is required.

Is my data leaving Atlassian when I use the CLI?

No. Teamwork Graph CLI is a local binary that runs on your machine. It calls Atlassian's cloud APIs to retrieve data, but your data is not stored or processed outside of Atlassian's infrastructure. Additionally, Teamwork Graph honours your IP and location allowlists if you have configured them.

What data does Teamwork Graph CLI access and who can see it?

Teamwork Graph CLI respects your existing Atlassian permissions. It only surfaces data that the authenticated user is already authorised to see. No new access is granted by installing or using the CLI.

Does Teamwork Graph CLI support OAuth?

Yes. OAuth 2.1 is the only supported authentication method, and interactive twg login uses OAuth by default. TWG CLI also supports delegated OAuth for embedding applications that own the credential lifecycle. Bitbucket is the one exception: it isn't covered by OAuth yet and requires a separate Bitbucket token.

Do Bitbucket commands require `--workspace` and `--repo`?

Commands that target a repository, including twg bb prs mergeability, require both a workspace and a repository to resolve. The flags can be omitted when another input provides that context. A PR number such as 42 identifies a PR within the selected repository.

Target input	Workspace and repository source
`--workspace my-workspace --repo my-repo`	Both values are explicit.
`--repo my-workspace/my-repo`	Both values come from the repository reference.
`--repo https://bitbucket.org/my-workspace/my-repo`	Both values come from the URL.
Neither flag, inside a Bitbucket checkout	The local CLI reads a Bitbucket Git remote, checking `origin` first, then the remaining remotes.

For example, replace the workspace, repository, and PR number with your target:

1
2
twg bb prs mergeability 42 --repo my-workspace/my-repo

Use an explicit target in CI, scripts, and agent calls that might run outside the intended checkout. Login and --site do not select a default Bitbucket workspace. Workspace operations such as twg bb workspace get need only a workspace.

For mergeability, Git detection currently runs only when both target flags are omitted. Passing a bare --repo my-repo alone does not fill in the workspace from Git; passing --workspace my-workspace alone does not fill in the repository. Use one of the complete targeting forms above.

Missing required context produces a local validation error before the product API request. An HTTP 404 after context resolves may indicate a missing or inaccessible target, or an unavailable endpoint; it does not prove that the workspace itself is missing.

Does Teamwork Graph CLI only work with agent terminals?

No. Teamwork Graph CLI can be used by any AI agent or tool that can invoke a CLI. For example, agents that run commands behind the scenes — not just agent terminals — can use Teamwork Graph CLI.

When using Confluence via the TWG CLI, which models are recommended for whiteboard and database generation?

When using twg confluence content create through the TWG CLI, we recommend some models for better whiteboards and database generation.

For whiteboards, Sonnet 4.6 is recommended. GPT 5.5 is compatible, but other models have not been tested for quality.

For databases, Sonnet 4.6 and Opus 4.6 are recommended.

Should I use Teamwork Graph CLI or Rovo MCP?

Teamwork Graph CLI and Rovo MCP are both official ways for AI agents to work with Atlassian data via the Teamwork Graph. They are complementary, not interchangeable:

TWG CLI is a command-line, agent-oriented interface best suited to shell, CI/CD, and deep, graph-style automation.
Rovo MCP is an MCP server that exposes tools to MCP-capable hosts (web LLMs, IDEs, or sandboxes) using OAuth and host-native auth flows.

For a detailed comparison and decision guide, see Choosing between Teamwork Graph CLI and Rovo MCP.