Skip to content

atlassian.md

Latest commit

聽

History

History
64 lines (50 loc) 路 2.9 KB

atlassian.md

File metadata and controls

64 lines (50 loc) 路 2.9 KB
Raw

Jira and Confluence

Back to the docs

ldo jira and ldo confluence read Atlassian Cloud as your own account, with an API token: they see what you can in Jira and Confluence, and nothing more.

ldo jira whoami                                   # who the token reads as, and on which site
ldo jira issues --project OPS                     # a project's issues that are not done
ldo jira issues "assignee = currentUser() AND statusCategory != Done ORDER BY priority DESC"
ldo jira issue OPS-123                            # one issue, its description as Markdown
ldo jira issue OPS-123 --markdown > OPS-123.md    # the description alone
ldo jira projects                                 # the projects you can see
ldo confluence spaces                             # the spaces you can see
ldo confluence pages --space OPS                  # a space's pages, newest change first
ldo confluence page 123456                        # one page, its body as Markdown
ldo confluence page 123456 --markdown > runbook.md
ldo confluence search 'type = page AND text ~ "restart"'

Signing in

Make an API token at id.atlassian.com, then either set three variables:

export JIRA_INSTANCE="https://contoso.atlassian.net"   # or just contoso
export JIRA_EMAIL="you@example.com"                    # the account the token belongs to
read -rsp "API token: " JIRA_TOKEN && export JIRA_TOKEN

or add profiles to the config file (ldo config init writes a template), one for each site, and pick one with -p (or LDO_ATLASSIAN_PROFILE):

[atlassian]
default_profile = "work"

[atlassian.profiles.work]
site = "https://contoso.atlassian.net"
email = "you@example.com"
# token_env = "JIRA_TOKEN"   the variable holding its token, for a second site

The token is never in the file, never on the command line, and never printed: it goes with the email as HTTP Basic, Atlassian Cloud's way for API tokens. Jira and Confluence share it.

Jira

issues takes a JQL query, or --project KEY for that project's issues that are not done. Jira refuses a query with no restriction at all, so without either it is every issue not done, newest change first; -n sets how many (50 by default). Each row has the key, type, status (green when done, amber in progress), priority, assignee, when it changed and the summary. issue shows one issue's details, then its description as Markdown, and --markdown prints the description alone.

Confluence

pages lists pages, the most recently changed first: every space's, or one's with --space (a personal space's key is ~ and an id, as spaces shows it), or those with one --title. page shows a page's details, then its body as Markdown: headings, emphasis, links, lists, tables, quotes, code (Confluence's code macro among it) and tasks, with macro settings left out. search takes CQL, Confluence's query language.