Fetch a Jira issue by key. Returns lean fields: key, summary, description (markdown), status, type, parent ({key, summary, type} or null), priority, assignee, reporter ({displayName, accountId} or null), labels, updated (ISO), updatedJql (updated floored to the minute in your Jira timezone, for JQL updated > "..." polling; at-least-once, dedupe; null if unavailable), comments (id, author, authorAccountId, body, created, updated), attachments (id, filename, mimeType, size, content URL). Attachment content URLs require Jira authentication.

Batch-read up to 50 Jira issues in one call (one JQL search, plus one GET per key the search omits). Returns an array in request order, one entry per unique key: {key, summary, type, status, parent ({key, summary, type} or null), description (markdown or null), comments: [{author, body, created}], commentsTotal?}. fields selects description and/or comments (default both). Descriptions are capped at maxDescriptionChars (default 600), comment bodies at maxCommentChars (default 300); cut text ends with "… [truncated, N chars total]". Only the 3 most recent comments are returned (oldest first); commentsTotal appears when more exist. A key that is missing or not visible to you returns {key, error} instead of being dropped. A moved/renamed key returns the issue under the requested key plus movedTo (current key). Use jcm_getIssue for full detail on one issue.

Get a Jira issue's field-change history, oldest first. Returns {entries: [{id, authorAccountId, author, created, items: [{field, from, to}]}], truncated}. since keeps entries created strictly after it. Capped at maxResults oldest matches (entries sharing the last timestamp stay together); if truncated, call again with since = last returned created.

Resolve an attachment ID from jcm_getIssue to {attachmentId, filename, url, temporary: true}. Requires the server's Jira account to have permission to view the attachment. Download the sensitive temporary URL without Jira credentials using a local download tool; request a fresh URL after expiry. File bytes are not returned. Supports Jira Cloud's Atlassian media redirects.

Get available workflow transitions for a Jira issue. Returns [{id, name}].

Update a Jira issue. Pass 'fields' as {fieldName: value} to set values (description accepts markdown), and/or 'update' for incremental ops like adding a fix version: update={fixVersions: [{add: {name: '1.2.0'}}]}. 'update' values are passed verbatim, not markdown-converted. At least one is required. Returns {ok, key}.

Move a Jira issue to a new workflow state. Use jcm_getTransitions first to get valid transition IDs. Returns {ok, key, transitionId}.

Convert one or more Jira issues (same project) to another issue type, keeping their keys. To make tasks sub-tasks, provide parentKey. To make sub-tasks standalone, omit parentKey. Uses Jira's bulk move and waits up to 30s. Returns {ok, keys, issueType, parentKey}, or {ok:false, taskId, status, failed?} if the move failed or is still running.

Add a Jira issue to its project's active scrum sprint. No-op if the issue is already in an active sprint or the project has no scrum board. If the board or sprint is ambiguous, returns {ok:false, reason, candidates:[{id,name}]}: for reason multiple-scrum-boards retry with boardId; no-active-sprint and multiple-active-sprints need manual handling in Jira. Returns {ok, key, sprintId, sprintName, boardId} or {ok, skipped}.

Returns the authenticated Atlassian user's account info: accountId, displayName, email, active.

Look up a Jira user by email address. Returns [{accountId, displayName, email, active}].

Create a new Jira issue. Returns {key, url}. For sub-tasks, pass parentKey.

Add a comment to a Jira issue. Body accepts markdown (headings, tables, code fences, lists). Returns {ok, commentId}.

Search Jira issues using JQL. Returns {total, issues: [{key, summary, status, type, parent, assignee, updated, updatedJql}]}. parent is {key, summary, type} or null. updatedJql is the issue's updated time floored to the minute in your Jira timezone; poll with key = X AND updated > "<updatedJql>" (at-least-once, dedupe). Zero results are validated: invalid JQL values (unknown issue type, missing parent, etc.) return an error instead of an empty list.

Create a new version (release) in a Jira project. Returns {id, name}.

List all versions in a Jira project. Returns [{id, name, released, releaseDate}].

Mark a Jira version as released. Returns {ok, id, name}.

Create a link between two Jira issues. issueKey is the source and gets the outward verb (e.g. "blocks", "duplicates", "clones"); targetIssueKey gets the inward verb. Returns {ok, issueKey, targetKey, linkType}.

Delete an issue link by its ID. Use jcm_getIssueLinks to find link IDs. Returns {ok, linkId}.

Get all links for a Jira issue. Returns flat array of [{id, type, inwardIssue?, outwardIssue?}].

Fetch a Confluence page by ID. Returns {id, title, body (markdown), status, version, spaceId, parentId?}.

Resolve a Confluence space key to its numeric ID for use with jcm_createArticle. Returns {id, key, name}.

Create a new Confluence page. Returns {id, title, url}.

Update a Confluence page. Pass current version number (auto-incremented). Returns {ok, id, version}.

Search Confluence pages using CQL. Returns [{id, title, space, spaceKey, lastModified}].