confluence-best-practiceslisted

# confluence-best-practices — use Confluence leanly, structure content sensibly, write it readably Primary reader: an **agent helping a user** — one that *acts on* Confluence (creating, organising, and governing content) and advises the user on **how to use Confluence well**: what a piece of content *is*, where it *goes*, how to keep the wiki lean and findable, and how to make pages people can actually read. This is the **judgment layer** above the execution mechanics — it decides *what* should change and *how content should be structured*; the tools do the *how*. Two audiences, in priority order: **(1) an agent acting automatically** on Confluence (the main case), and **(2) a human authoring content** who wants it readable. The good news is these converge — the structure that makes a page scannable for a person is the same structure that makes it retrievable for an agent. This skill is **instance-agnostic and organisation-adaptive**. It never assumes a team's space taxonomy, conventions, or working language — it shows how to *discover* the org's actual conventions and reason within them. It is **self-hosted-first**: defaults are for **Confluence Data Center 10.2 / 9.2 LTS**, but everything has a **Cloud** counterpart noted inline (see the Cloud-vs-DC guard). ## How the work gets done (execution layer) An agent here acts through **content-level REST/CQL** — directly, or via an MCP server (e.g. `mcp-atlassian`), or by handing the user UI steps. The boundary that determine