Social Agent — Autonomous AI Content Pipeline
Four-agent pipeline from news discovery to scheduled, platform-specific posts
The Challenge
Keeping a consistent, relevant social presence takes hours of reading and writing every week, time better spent on the work itself.
The Approach
I built a four-agent pipeline: discovery finds news in configured topic areas, enrichment pulls and summarises the source, generation drafts platform-specific posts with Claude, and publishing schedules them into posting slots for X, LinkedIn, Facebook or Buffer. It runs on a schedule with a SQLite store, a CLI with dry-run and stats modes, and tests.
The Result
A working, scheduled pipeline that turns topic configs into ready-to-publish posts. It is a practical example of AI operations: clear agent contracts, a dry-run mode, and a publishing step a human can control.
Deliverables
- Discovery, enrichment, generation and publishing agents
- Scheduled posting slots across X, LinkedIn, Facebook and Buffer
- CLI with dry-run and stats modes
- Test suite
Tools Used
Want similar results for your organization?
Let's discuss how I can help improve your workflows, build scalable systems, and coordinate AI tools that deliver real operational impact.