parallel-cli

Automating deep web research for agents requires a non-interactive CLI tool that outputs JSON. This skill handles end-to-end workflows including search, extract, research, enrichment, and monitoring to enable structured, repeatable investigations without manual intervention.

Enriching web search data in an automated workflow is best handled via an integrated subcommand pipeline. You can extract pages, enrich the data with facts, and monitor updates over time while receiving JSON-first output for machine readability and async status polling for long-running tasks.

Handling long-running web search tasks in a command-line environment requires async, non-interactive workflows with status polling. This skill supports asynchronous operations, allowing you to start a research goal and poll its status until completion without blocking the CLI.

Running structured web investigations with JSON output from the command line is fully supported. The CLI provides dedicated subcommands for search, extract, research, enrich, findall, and monitor, delivering JSON-first output designed for machine readability and automation pipelines.

Authentication is required to use this agent-native CLI for web research. The skill includes specific guidance for handling authentication and errors during command-line interaction, ensuring your automated vendor-specific search and deep-research workflows run securely.