confluence_skills

Search, browse, retrieve, cache, and summarize Confluence spaces, pages, page hierarchies, comments, and attachments through a local Python helper script. For normal page-search tasks, use the views-ranked two-stage workflow by default: recall up to 50 relevance results and 50 recent results, fetch views for candidates, then read the top 5 by views. Use when Codex needs to search Confluence documentation, read a page, list spaces/pages/children, inspect attachments, download content, answer questions grounded in Confluence pages, or cache retrieved Confluence content locally. 适用于用户要求“查 Confluence 文档”“读取 wiki 页面”“在 cf 上查”“去 CF/Confluence 查”“搜索空间或页面”“在某个疑似 Confluence 空间、页面名称、wiki 名称或文档标题里查询”“获取页面内容”“列出子页面或附件”“基于 Confluence 回答问题”等场景。Do not use for deleting pages/comments, changing permissions, modifying space settings, or any administrative Confluence operation.

Generate enterprise-ready Markdown documentation from the current conversation context. Use when Codex needs to summarize the current session, convert chat context into work records, meeting notes, implementation records, project summaries, technical decision records, issue retrospectives, exception handling notes, handoff documents, Confluence-ready pages, or other structured documents based on the local/current conversation context. 适用于用户要求“总结当前会话”“生成工作记录”“整理实施过程”“输出会议纪要”“沉淀 Confluence 文档”“形成交接文档”“复盘问题处理过程”等场景。

Search, browse, retrieve, cache, and summarize Jira issues, projects, boards, sprints, comments, attachments, users, and worklogs through a local Python helper script. Use when Codex needs to search Jira with JQL, read issue details, inspect project metadata, list comments or attachments, download attachment files, review agile board or sprint data, answer questions grounded in Jira items, or cache retrieved Jira content locally. Do not use for deleting issues/comments, changing permissions, modifying workflows, or administrative Jira operations.

Upload generated Markdown documentation to Confluence by creating new pages under a configured root page and username child page. Use when Codex needs to publish the latest context-transform cache document, upload a local Markdown file, create a Confluence page, or ensure the configured username child page exists under root_page. Works as the publishing step after context-transform generates a Markdown document in context-transform/cache.