Every week you have AI conversations worth keeping—a debugging session that finally clicked, a research thread surfacing the right papers, a design argument you don't want buried in chat scroll. Then the tab closes and your model forgets everything by next Monday. blogs.city exists to close that loop for good.

What Is blogs.city Actually?

blogs.city blogifies and wiki-fies the dialogue you already have with large models. Register, connect your AI editor—Claude Code, Cursor, Trae, OpenClaw, or any editor speaking Skills—and publish high-value exchanges directly as posts, wiki pages, and durable body-text memory your model can read back on the next turn. The pipeline is simple: You + LLM talk → publish the outcome → blog + wiki → next session reads the body. No copy-paste cleanup, no manual archiving. The conversation becomes the permanent record.

Hermes Integration: The Real Win

This isn't another chat wrapper—it's infrastructure that makes LLM dialogue compound. For Hermes users, blogs.city provides a direct fit since Hermes already organizes local LLM wikis around ingest, query, and lint conventions. blogs.city exports the same shape into your live blog: SCHEMA.md becomes auto-generated under wiki/SCHEMA.md, index.md maps to your entities/concepts/comparisons/queries, and log.md tracks activity. The publish templates align perfectly with Hermes's mental model—one discipline, two surfaces, aligned semantics. OpenClaw and other AI editors connect via personal API token and the publish Skill.

Page Types: Blogification Plus Wiki-fication

Not every chat message deserves a post. The ones that do get structured roles: Source pages capture digests of external papers or docs; Concept pages track terms your threads keep returning to; Entity pages stabilize references for people, products, or tools; Synthesis pages document queries answered across multiple prior posts; Article pages handle everything else—tutorials, notes, field reports. Categories tell you the topic domain while page type tells you which wiki job applies. Links live inside prose where they carry meaning, not in footer blocks.

Body-Text Memory: The Actual Breakthrough

Chat memory is shallow—truncated, session-bound, often wrong. blogs.city takes a different approach with body-text memory: every published post is plain Markdown on disk. Your AI editor can call the catalog to see what concepts, syntheses, and entities already exist before writing anything new. Search pulls candidates by keyword; Read loads full body text into context. The valuable exchange you had last month isn't a faded chat log—it's paragraphs your model can quote, link, and contradict on purpose. Your blog becomes your LLM's external memory: auditable, linkable, yours.

Built-In Blog Infrastructure

You get a real blog from day one—posts, nested categories, tags, archives, comments, RSS, five built-in themes, interactive charts in Markdown, mobile layout with infinite scroll, favorites and reading history. SEO fields populate on publish. AI-assisted comment screening keeps noise down. This isn't a minimal wrapper—it's production blogging infrastructure with the AI wiring where it matters: at publish, discover, link, and remember.

Your Files Stay Yours

Markdown files live on disk. Redis handles index speed. Backup and offsite sync come standard. Export to Obsidian Vault ZIP or single-post PDF whenever you want portability. Wiki views regenerate from the catalog—never a second source of truth. The platform runs the house; you keep the keys and the corpus.

Key Takeaways

  • blogs.city transforms AI conversations into searchable Markdown posts without manual cleanup
  • Hermes integration aligns perfectly with existing wiki conventions for dual-surface knowledge management
  • Body-text memory lets your AI editor read prior exchanges as quotable, linkable paragraphs
  • Full blogging infrastructure (RSS, themes, comments) ships alongside the AI memory features