菜单
Use when reading or writing Jira tickets, fetching Confluence pages, searching with JQL, adding comments, or downloading Atlassian attachments from this devbox.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use on any NixOS host (devbox on Hetzner, cloudbox on GCP) when SSH connection fails, host key mismatch, NixOS issues, CPU/IO contention (high load), or verifying the host is properly configured
Launch headless opencode sessions from CLI. Use when you need to start a new opencode session in the background to work on a task in parallel, or when spawning work on a specific directory.
How to apply configuration changes to NixOS hosts (devbox, cloudbox). Use when you need to rebuild the system, apply home-manager changes, or recover from issues.
Use when setting up or rotating PagerDuty MCP auth for OpenCode on macOS Keychain or cloudbox sops, or when debugging missing PagerDuty tools.
Use when setting up or rotating Rollbar MCP auth for OpenCode on macOS Keychain or cloudbox sops, when triaging a Rollbar error paged via PagerDuty, or when debugging missing Rollbar tools.
Documents the OpenCode agent set — what each does, when to use it, and why the others were cut. Use when questioning agent choices or considering adding/removing agents.
| name | using-atlassian |
| description | Use when reading or writing Jira tickets, fetching Confluence pages, searching with JQL, adding comments, or downloading Atlassian attachments from this devbox. |
Two complementary tools. Read this whole file before choosing one — picking the wrong tool is the most common mistake.
atlassian MCP server): mutations, JQL/CQL search, arbitrary fields, and Confluence create/edit. This is the default for everything except rich reads.:FetchJiraTicket / :FetchConfluencePage): the best read path — it renders a ticket or page (plus threaded comments and image attachments) to Markdown on disk.| Workflow | Use |
|---|---|
| Read a Jira ticket as context (rendered Markdown + image attachments) | nvim :FetchJiraTicket |
| Read a Confluence page (with comments + PNG diagrams) | nvim :FetchConfluencePage |
| Archive Atlassian content to disk for offline reference | nvim |
| Create a new Jira ticket | MCP createJiraIssue |
| Edit a ticket's summary / description / status / assignee | MCP editJiraIssue / transitionJiraIssue |
| Add a comment | MCP addCommentToJiraIssue |
| JQL search ("my open tickets in epic X") | MCP searchJiraIssuesUsingJql |
| Read non-default Jira fields (custom fields, labels, story points, sprint, fix version, parent epic) | MCP getJiraIssue |
| Confluence page body only, no images/comments | Either; confluence-to-md.sh is lightest |
| Confluence with comments or attachments | nvim (only option) |
| Create / edit Confluence pages or comments | MCP createConfluencePage / updateConfluencePage |
Rule of thumb:
nvim --headless ticket.md -c "FetchJiraTicket PROJ-1234" -c "write" -c "quit"
# → ticket.md (Markdown), images in ~/.cache/atlassian-attachments/jira/PROJ-1234/
nvim --headless out.md -c "FetchConfluencePage 1234567890" -c "write" -c "quit"
# → out.md (Markdown + threaded comments), PNGs in ~/.cache/atlassian-attachments/confluence/1234567890/
# Extract page ID from URL:
echo "https://company.atlassian.net/wiki/spaces/ENG/pages/1234567890/Title" | grep -oE '[0-9]{10}'
Call the atlassian MCP's createJiraIssue tool with cloudId, projectKey
(PROJ), issueTypeName (Task, Bug, Story, Epic, …), summary, and a
Markdown description. Set the assignee to $ATLASSIAN_ASSIGNEE_ID.
Format gotcha — always use GitHub-Flavored Markdown, never Jira wiki markup. The Atlassian MCP (and the underlying Cloud REST API v3) expects descriptions and comment bodies in Markdown — # heading, **bold**, `code`, [text](url), | col | col | tables — and converts to ADF on the way in. Jira wiki markup (h2., *bold*, {{code}}, [text|url], || col || col ||) renders as literal text and is the most common mistake when an LLM drafts a ticket. If you see your description rendering with literal h2. or {{...}} in the Jira UI, you used wiki markup — re-edit with Markdown.
Call addCommentToJiraIssue with the issueIdOrKey (PROJ-1234) and a Markdown commentBody. Comment bodies follow the same rule: Markdown only, never wiki markup.
Call searchJiraIssuesUsingJql with a jql string, e.g.
project=PROJ AND assignee="<accountId>" AND statusCategory!=Done.
Call getJiraIssue with the issueIdOrKey. It returns the full field set (custom fields, labels, story points, sprint, fix version, parent epic). For rendered Markdown of the description + comments + image attachments, use nvim instead.
| Setting | Value |
|---|---|
| Subdomain | $ATLASSIAN_SITE |
| Project | PROJ |
| Default Assignee | $ATLASSIAN_ASSIGNEE_ID |
| Active Epic | PROJ-5678 |
Create pages with createConfluencePage and edit them with updateConfluencePage (pass cloudId, spaceId/pageId, title, and Markdown body). Footer/inline comments are also available via the MCP's Confluence comment tools. For reading a page with its threaded comments and diagrams rendered, still prefer nvim :FetchConfluencePage.
See REFERENCE.md for:
confluence-to-md.sh body-only helperThe confluence-to-md.sh helper script lives alongside this skill.