What problem does it solve?
Autoresearch removes the burden of manually chaining research, debugging, verification, and iteration by turning a goal into a controlled autonomous loop that keeps improving until a measurable outcome is reached.
Core Features & Use Cases
- Goal-to-loop orchestration: Converts a plain objective into a routed workflow with plan, debug, fix, security, ship, or research-style subcommands.
- Metric-driven iteration: Repeats modify-verify cycles against explicit success predicates, keeping or discarding changes based on evidence rather than guesswork.
- Safety and convergence controls: Adds bounded iterations, handoff files, regression checks, independent verification, and stop conditions to prevent runaway execution.
- Use case: Use it when you want an AI coding agent to debug a failing system, harden a feature, or improve a metric without manually supervising every step.
Quick Start
Use the autoresearch skill to turn my goal into the right iterative workflow and run it with safe, metric-based convergence.
