autoresearch

Persist research loops until a completion artifact passes validator approval.

Updated Apr 20, 2026
One-click install
npx skills add https://github.com/hotaq/Sprite_harmess --skill autoresearch-hotaq
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: autoresearch
Source: https://github.com/hotaq/Sprite_harmess/tree/main/.codex/skills/autoresearch
Command: npx skills add https://github.com/hotaq/Sprite_harmess --skill autoresearch-hotaq

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill replaces ad hoc, stop-when-it-feels-done research loops with a persistent workflow that keeps nudging toward explicit validation evidence before finishing.

Core Features & Use Cases

  • Validation-Gated Execution: Choose a validation regime up front and let the loop continue until the required artifact proves success.
  • Stateful Persistence: Store mode state, completion targets, and output references so the workflow can recover cleanly between runs.
  • Use Case: Use it when you need a research task to keep iterating on findings, evidence, and outputs until a validator or architect approves the result.

Quick Start

Use the autoresearch skill to pick a validation mode, persist the state files, and continue the research loop until the completion artifact passes validation.

Frequently Asked Questions about autoresearch

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

FAQPage Schema
How do I keep a research loop running until validation evidence is found?

A persistent research loop applies repeated nudges to ongoing investigations and enforces artifact-gated completion, ensuring explicit validation evidence is produced before the workflow stops.

What is artifact-gated completion for long-running research tasks?

Artifact-gated completion is a validation mechanism that requires a completion artifact to pass explicit validation before a long-running research task is allowed to stop, ensuring findings are verified by a validator or architect.

How do I recover state for a research workflow between runs?

You can recover research workflow state between runs by persisting mode state, completion targets, and output references, allowing the investigation to resume cleanly without losing progress from previous iterations.

When do I need stateful persistence for research investigations?

You need stateful persistence for research investigations when tasks are long-running and require repeated nudges, state recovery between runs, and explicit evidence before stopping, ensuring the workflow validates findings before finishing.

Does this persistent research workflow require validator or architect approval to finish?

Yes, the persistent research workflow requires validator or architect approval to finish, enforcing artifact-gated completion so the loop continues iterating until the required completion artifact passes the chosen validation mode.

What is the best way to manage long-running investigation tasks that need repeated nudges?

The best way to manage long-running investigation tasks needing repeated nudges is a persistent workflow that stores mode state and enforces artifact-gated completion, continuing iterations until validation evidence is explicitly approved.