autoresearch

Orchestrate a persistent, validator-gated research loop with state storage.

6|1|Updated Apr 16, 2026
One-click install
npx skills add https://github.com/Goblin1024/oh-my-kimi --skill autoresearch-goblin1024
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: autoresearch
Source: https://github.com/Goblin1024/oh-my-kimi/tree/main/skills/autoresearch
Command: npx skills add https://github.com/Goblin1024/oh-my-kimi --skill autoresearch-goblin1024

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Orchestrates a persistent, validator-gated research loop within a native-hook stateful workflow, addressing the need for continuous evaluation and artifact-backed conclusions.

Core Features & Use Cases

  • Persistent research sessions with gated validation modes (script or architect-based).
  • Structured state persistence and artifact-driven completion for auditability.
  • Seamless workflow coordination between intake (deep-interview) and execution (autoresearch).

Quick Start

Run a deep-interview to clarify mission and evaluator, then start autoresearch with the chosen validation mode.

Frequently Asked Questions about autoresearch

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I manage persistent research sessions with iterative validation?

Persistent research sessions are managed through a validator-gated research loop that stores state continuously. This enables iterative validation by retaining session context and progress across long-running tasks until a passing validator completes the workflow.

What is a validator-gated research workflow and when do I need it?

A validator-gated research workflow is an orchestration loop that requires continuous evaluation before advancing. You need it for long-running research tasks demanding iterative validation and artifact-backed conclusions for strict auditability.

How do I choose between script-based and prompt-architect evaluation for research tasks?

You choose between script-based and prompt-architect evaluation during the init-time validation-mode selection. This configures the research loop to apply either automated script validation or architect-approved artifact evaluation for completion.

Can I use stateful workflow orchestration for long-running research without losing progress?

Yes, stateful workflow orchestration persists progress automatically to a local state file. This ensures long-running research tasks maintain their state and validation context across interruptions until a final artifact is approved.

How do I start a research workflow using a deep-interview?

You start by running a deep-interview to clarify your mission and select an evaluator. The intake process then seamlessly coordinates with the research execution loop using your chosen validation mode.

What are the limitations of artifact-based completion in research workflows?

Artifact-based completion requires either a passing script validator or an architect-approved artifact. The research loop will not finalize conclusions until this strict validation gate is satisfied, potentially extending task duration.