Orchestrate end-to-end research sweeps from goal to execution.
npx skills add https://github.com/Borda/AI-Rig --skill sweep-borda
Please help me install this Agent Skill. Skill: sweep Source: https://github.com/Borda/AI-Rig/tree/main/plugins/research/skills/sweep Command: npx skills add https://github.com/Borda/AI-Rig --skill sweep-borda
Non-interactive end-to-end research pipeline: auto-plan → judge gate → run. Single command from goal to result.
Invoke the sweep command with a quoted goal to start the end-to-end pipeline.
High-intent search queries and answers about installing and using this skill.
Automating end-to-end research sweeps requires a tool that plans, judges, and runs tasks from a single goal. This Skill orchestrates the entire pipeline non-interactively, applying a judge and refine loop before final execution.
A judge and refine loop in automated research execution evaluates a generated plan before running it. This Skill enforces up to three iterations of this loop to ensure plan quality and passes safety checks before final run orchestration.
Running non-interactive research pipelines with team collaboration is supported through specific flags. You can invoke the command with the --team flag to enable team-assisted runs during the orchestration process.
Configuring compute options for automated research sweeps is supported within the execution phase. You can apply flags like --colab to adjust the compute environment during the final run orchestration.
The judge and refine loop in research automation is limited to a maximum of three iterations. This constraint ensures quality control before running while preventing endless refinement cycles during the planning phase.
Sweep is designed for non-interactive research orchestration, automating the plan, judge, and run steps from a single command. It requires no manual intervention once the goal is provided to the pipeline.
vikingokftCreate a Google Tasks list and seed it with initial tasks.
Ocean326Poll a research task by run ID and retrieve its result in JSON.
lucaswallOrchestrate an agent team to implement tasks from PLANS.md with progress tracking and verification.
cbbkrd-techOrchestrate multi-phase test planning by delegating research, manual testing, and automated test planning tasks.
