Blog¶

The blog pipeline synthesizes weekly and thematic blog posts from your journal entries and reading digests.

Basic usage¶

uv run python -m distill blog --output ./insights

Post types¶

Weekly synthesis¶

Summarizes a week of journal entries into a cohesive blog post.

uv run python -m distill blog --output ./insights --type weekly --week 2026-W09

Thematic deep-dives¶

Detects recurring themes across journal entries and generates focused posts.

uv run python -m distill blog --output ./insights --type thematic

Configuration¶

[blog]
target_word_count = 1200
include_diagrams = true        # Mermaid diagrams
model = "claude-sonnet-4-6"
platforms = ["obsidian"]       # where to publish

Publishing¶

Obsidian (default)¶

Blog posts are written to ./insights/blog/ as markdown with YAML frontmatter and Obsidian wiki links.

Ghost CMS¶

uv run python -m distill blog --output ./insights \
  --publish ghost \
  --ghost-url https://your-blog.ghost.io \
  --ghost-key "id:secret"

See Ghost integration.

Multiple platforms¶

uv run python -m distill blog --output ./insights --publish obsidian,ghost,postiz

Memory¶

The blog pipeline tracks what it has generated (.blog-state.json) to avoid regenerating existing posts. It also maintains cross-post memory (.blog-memory.json) to avoid repeating examples.

Editorial notes¶

Guide the blog pipeline with editorial notes:

uv run python -m distill note "Focus on the multi-agent architecture this week" \
  --target "week:2026-W09"

Notes are injected into the LLM prompt for the specified target.