parallel-cli

Automated web search workflows are orchestrated using a CLI that delivers structured JSON outputs for extraction and research tasks. This approach replaces bespoke code by enabling non-interactive, JSON-driven pipelines for scalable information discovery.

Long-running research and monitoring tasks are managed asynchronously using status and poll workflows. The CLI supports async job execution with context chaining via previous-interaction-id, allowing agents to track and resume multi-step operations.

Data enrichment and monitoring are integrated within a single CLI, supporting structured discovery across multiple sources. It handles async and one-shot modes, enabling enterprise-grade information discovery and enrichment without requiring bespoke orchestration code.

Structured JSON from web research tasks is returned using non-interactive CLI flows with explicit arguments. This JSON-ready output format supports scalable research pipelines, deep research, and FindAll operations directly from the command line.

Multi-step workflows are supported through context chaining via previous-interaction-id. This allows sequential research and enrichment tasks to maintain state across interactions, enabling complex, multi-stage web research pipelines.

Non-interactive CLI web research requires explicit arguments and JSON output, limiting real-time interactive adjustments. It is designed for structured, scalable pipelines rather than exploratory, ad-hoc browsing, relying on status and poll mechanisms for long-running tasks.