How to Write a Curated Newsletter with Logseq

A curated newsletter workflow in Logseq can be built end-to-end around frictionless capture, query-based retrieval, and clean exporting—turning scattered links and notes into a repeatable weekly (or project-based) publication. The core idea is to treat newsletter writing as a pipeline: collect resources on the Journals page using a template, maintain an always-up-to-date index via Logseq queries, then assemble a workbench page that becomes the draft for publishing. The payoff is consistency: resources don’t get lost, previously covered items drop out automatically once marked “done,” and the newsletter can be produced quickly without manual sorting.

The process starts with defining what “curated” means in practice: not just finished articles, but links, highlights, meeting notes, and action points—anything worth sharing in a distilled format. Ramses frames the workflow as useful for both personal learning newsletters and corporate stakeholder updates, including communities of practice where experts share tools and tips. Logseq is chosen because it makes entry low-friction (templates on the Journals page), preserves structure (templates enforce consistent tags/properties), and supports retrieval (queries) and export (Markdown or HTML).

Collection hinges on a template created on a dedicated Templates page. In the demo, each newsletter resource is captured as a to-do block with a consistent set of identifiers: an area tag (e.g., “#newsletter” under “area/Logseq”), a namespace-like area (such as “area/Logseq”), and a resource marker (e.g., “resource”). The template also includes a “to-do” property/state so the system can track whether an item has already been included. Because entries are made on the Journals page, Logseq’s date-aware query filters (like “between 7 days ago and today”) can later narrow results to the right time window.

Retrieval is handled by a query that functions like a dynamic index. The query searches for blocks matching the template’s identifiers (area/Logseq, #newsletter, and “resource”) and then filters by the to-do state. Items marked “done” disappear from the query results, preventing repeated newsletter mentions. A second filter using the “between” operator limits output to a specific period—useful for weekly action points or meeting summaries—so the newsletter draft can be generated from only the relevant slice of captured material.

Drafting happens on a dedicated workbench page (e.g., an “area/Logseq/newsletter” page) where the query results are dragged or copied in. Each newsletter section becomes a heading, with the associated resources nested beneath it. From there, exporting is done by copying a heading or block as Markdown (“copy as”) and pasting into a publishing platform. The demo uses Ghost for publishing because it supports newsletter sending directly to subscribers, whereas Logseq Publish lacks built-in newsletter functionality. When Markdown export produces bullet formatting that’s inconvenient, the workflow includes quick cleanup (or alternative export approaches like HTML for email clients). The session also touches on tooling options to improve presentation (e.g., Hugo export via Logseq plugins, PDF export plugins, and document-mode HTML exporters) and ends with discussion of Logseq’s future direction, including performance improvements and potential AI-assisted features like natural-language query support and research question generation via elicit.