generate-content

To generate social media drafts, provide a specific idea_id to read the kept idea from the database, run a research pass, and produce platform-native posts for TikTok, Instagram, and LinkedIn. The system outputs ready-to-post drafts for your review.

No, the content generation pipeline does not schedule posts automatically. It produces and gates drafts for an explicit approval step, ensuring nothing gets scheduled or published without your manual review.

The writer and critic loop works by using separate agents per platform to draft and critique content. The writer generates the post, the critic scores it, and the system iterates until producing critic-approved drafts saved in the database.

Yes, you can generate TikTok and LinkedIn posts with distinct voice profiles. The system creates drafts across TikTok EN, TikTok DE, Instagram, and LinkedIn, applying the correct voice profile for each specific platform.

When research data is thin during the draft generation process, the system flags the gap within the cached research brief. This caching mechanism allows you to identify and re-run the research pass when more data is needed.

You need a valid idea_id stored in the data/content.db database to start generating drafts. The system reads the idea from the database, executes orchestrator agents per platform, and renders deterministic slide assets for approval.