From atlassian-suite
This skill should be used when the user asks to "init atlassian", "initialize atlassian", "setup atlassian", "configure atlassian", "configure jira/confluence/bitbucket", "set my atlassian credentials", "save atlassian credentials", "what credentials does the atlassian plugin need", "rotate my atlassian token", "remove atlassian credentials", or runs `/atlassian-suite:init`. Walks the user through configuring credentials for the Acendas Atlassian Suite — file-based by default, env-var fallback for CI.
npx claudepluginhub acendas/acendas-marketplace --plugin atlassian-suiteThis skill is limited to using the following tools:
Configure credentials for the unified Acendas Atlassian Suite MCP server (Jira Cloud + Confluence Cloud + Bitbucket Cloud). Default storage is `~/.acendas-atlassian/config.json` (mode 0600, owner-only, atomic-write with rolling backup). Env vars override the file when set.
Creates isolated Git worktrees for feature branches with prioritized directory selection, gitignore safety checks, auto project setup for Node/Python/Rust/Go, and baseline verification.
Executes implementation plans in current session by dispatching fresh subagents per independent task, with two-stage reviews: spec compliance then code quality.
Dispatches parallel agents to independently tackle 2+ tasks like separate test failures or subsystems without shared state or dependencies.
Configure credentials for the unified Acendas Atlassian Suite MCP server (Jira Cloud + Confluence Cloud + Bitbucket Cloud). Default storage is ~/.acendas-atlassian/config.json (mode 0600, owner-only, atomic-write with rolling backup). Env vars override the file when set.
Check current state. Call mcp__acendas-atlassian__get_credentials_status. Report:
ATAT...xyz9 (192 chars))env:JIRA_USERNAME, env:ATLASSIAN_USERNAME, or file)Decide the path:
configure_credentials with just atlassian_api_token (or per-product if they use different tokens). Existing fields are preserved.clear_credentials with confirm: true. Tell user env vars are unaffected.Collect what's needed — IMPORTANT: Jira/Confluence and Bitbucket use SEPARATE credentials.
Explain this to the user upfront:
Atlassian API tokens (from https://id.atlassian.com/manage-profile/security/api-tokens) authorize Jira + Confluence on a given Atlassian site. Bitbucket Cloud uses separate credentials — typically an app password, a Repository/Project/Workspace Access Token, or an Atlassian API token tied to a Bitbucket-enabled account. The two are usually scoped to different accounts and must be set separately.
Ask per product. Skip a section if the user isn't configuring it this run.
For Jira + Confluence (shared — same Atlassian site + same account + same token):
https://ventek.atlassian.net. Confluence URL is typically <base>/wiki.For Bitbucket (always separate):
bitbucket.org/<workspace>/...).Persist. Call mcp__acendas-atlassian__configure_credentials with fields scoped to the product they belong to:
For Jira + Confluence — use the shared atlassian.* fields (they fan out):
atlassian_username (the Atlassian-site email)atlassian_api_token (the Atlassian API token)jira_url, confluence_urlFor Bitbucket — use per-product bitbucket.* fields (so it doesn't accidentally get used for Jira/Confluence):
bitbucket_workspacebitbucket_usernamebitbucket_api_tokenOptionally: jira_projects_filter, confluence_spaces_filter to scope tools.
Do NOT put the Bitbucket token in atlassian_api_token unless the user explicitly confirms that ONE token works for all three products (rare — typically only when one Atlassian account has access to Jira + Confluence + a Bitbucket-linked workspace).
The tool merges into the existing config: values you don't pass are preserved. A backup of the prior file is kept at ~/.acendas-atlassian/config.json.bak. The tool's response includes a changes block showing exactly which fields were added, updated, or preserved — read this back to the user as confirmation.
Restart Claude Code. The MCP server only reads credentials at startup. Tell the user to restart Claude Code (or just the MCP server) for new credentials to take effect.
Verify. After restart, call mcp__acendas-atlassian__get_credentials_status again to confirm everything resolves. Run a sanity probe per configured product:
mcp__acendas-atlassian__jira_search with JQL assignee = currentUser() (limit 1)mcp__acendas-atlassian__getConfluenceSpaces (limit 5)mcp__acendas-atlassian__list_repositories (pagelen 5) — or use setup_bitbucket for an explicit credential pingIf any probe returns 401, walk through token rotation (back to step 4 with just atlassian_api_token).
For every credential value:
JIRA_USERNAME)ATLASSIAN_USERNAME)jira.username in config.json)atlassian.username)bitbucket_username + bitbucket_api_token explicitly. Don't rely on atlassian.* fallback for Bitbucket unless the user confirms one identity works for all three.~/.acendas-atlassian/config.json, mode 0600 (owner read/write only).~/.acendas-atlassian/config.json.bak — refreshed on every configure_credentials call. Recover from a mistake with mv config.json.bak config.json.jira_projects_filter or confluence_spaces_filter in the file to scope all tools to specific projects/spaces.READ_ONLY_MODE=true env var to disable all write tools.configure_credentials NEVER clobbers existing values — empty/undefined inputs are ignored. Verify via the changes.preserved field in the response.