newsletter-bluesky

Automating a Bluesky thread from monthly activity involves loading cached JSON summary data and generating a 3-4 post thread with specific constraints. This skill drafts plain text posts, enforcing one URL per post and at most two hashtags, saving the result to a markdown file.

Formatting automated Bluesky posts requires plain text output with zero markdown links or images. This generation process produces plain text posts, limits each post to one URL, and restricts the entire thread to a maximum of two hashtags for compliance.

Scoping social media content generation to a specific date range is supported using START_DATE and END_DATE inputs. Alternatively, providing a MONTH parameter constrains the content, loading cached activity data from the corresponding .cache/YYYY-MM/ directory.

Cached JSON files are required to generate a monthly social media summary. The process loads data directly from .cache/YYYY-MM/summary.json and per-repo JSON files to extract repository names and activity highlights for the thread.

Saving automated Bluesky drafts for publication workflows requires writing the final output to a structured file path. The generated thread is saved directly to newsletter/YYYY-MM/bluesky.md, ensuring the draft is ready for downstream publishing.

Structural limitations for generated Bluesky threads include a strict 3-4 post structure, a maximum of one URL per post, and no markdown links or images. The thread is limited to two total hashtags and references repositories in natural language.