Polywise is the open source agentic content system. It helps you capture content, build reusable context, and run agents against that context instead of starting from an empty chat every time.

You can save content, organize it by project or agent, search it through the retrieval pipeline, and then run sessions, tasks, or scheduled jobs that use that context.

No. Writing is one use case, but the system is built for broader agentic work: research, task execution, workflow coordination, link capture, and long-running project context.

Not always. What you need depends on which models or providers you want to run. Polywise is designed to work with different providers rather than forcing one billing path.

Yes. Polywise is built around provider flexibility, so you can connect the model stack that already fits your workflow instead of switching everything to a bundled vendor.

Yes. Polywise includes support for local runtimes such as llama-based setups, so part of the system can stay fully on your machine when you want tighter control.

Polywise uses a hybrid retrieval pipeline. It combines keywords, vectors, recall, and reranking so agents can recover relevant context instead of relying on a single search method.

Polywise is open source. Your content, memory, and runtime state remain inspectable and under your control, which makes backups, inspection, and privacy boundaries much clearer than cloud-only systems.

Yes. Polywise includes sessions, todos, schedules, link fetch jobs, and connector surfaces that can be composed into recurring workflows instead of manual one-off runs.

Yes. Polywise is available on GitHub, which means you can inspect the code, track changes, and understand the system you are building your workflow on.