This page lists every tool exposed by the Docsbook MCP server at https://docsbook.io/api/mcp/server. Each tool requires Bearer authentication via OAuth 2.0 + PKCE. The minimum plan column indicates the workspace tier required to call the tool; calls below that tier return a plan_required error.
Full-text/regex/heading/path search over the workspace's documentation content. Read-only — works with any token regardless of read/write scope.
get_doc_outline
Free
List every markdown page's title, heading count, and size before searching or writing. Read-only — works with any token regardless of read/write scope.
write_docs
Free
Commit one or more markdown files to the workspace's docs repo in a single atomic git commit. Requires a token authorized with read-write scope — a read-only token is refused.
For deeper local graph navigation (outline, fuzzy headings, link references, resolve links) while an agent has your docs checked out on disk, use markdown-lsp instead — run npx markdown-lsp <subcommand> ./docs to expose LSP-style doc_* tools on the working tree. See the markdown-lsp README for setup. search_docs/write_docs and markdown-lsp are complementary: the former work over the hosted MCP connection with no local checkout, the latter needs the repo on disk.
Registering a webhook is a Business-exclusive capability — Free and Pro/Pro+ workspaces cannot register any webhook regardless of event type.
Tool
Description
register_webhook_<event>
Register a webhook for one of ~15 events (HMAC secret + URL). Business only.
list_webhooks
List registered webhooks for the workspace
unregister_webhook
Remove a webhook subscription
list_webhook_deliveries
Delivery history with status, retry count, payload
replay_webhook_delivery
Re-deliver a specific past delivery
test_webhook
Send a synthetic payload to a URL
Event types include content.indexed, translation.completed, chat.no_answer, chat.negative_feedback, plan.upgraded, usage.limit_approaching, and others — see Webhooks for the full list and payload schemas.
Search the docs-skills catalog by query with optional category and requires_plan filters. Returns raw_url for the agent to fetch the SKILL.md directly.
