autoresearch

Automate a stateful validator-gated research loop with persistent .omx/state storage.

Updated Jan 4, 2026

Search Tags:#automation #validation #ai-research #workflow-management #autoresearch #stateful-workflow #native-hook

Ready to run

npx skills add https://github.com/YeaChan05/Remittance --skill autoresearch-yeachan05

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: autoresearch
Source: https://github.com/YeaChan05/Remittance/tree/main/.codex/skills/autoresearch
Command: npx skills add https://github.com/YeaChan05/Remittance --skill autoresearch-yeachan05

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Autoresearch replaces deprecated manual evaluation surfaces by providing a stateful, validator-gated loop that persists progress and results across cycles.

Core Features & Use Cases

Stateful loop orchestration with selectable validation modes (mission-script vs prompt-architect)
Artifact persistence under a dedicated .omx/state path for auditability
Safe, repeatable evaluation workflows suitable for AI validation tasks

Quick Start

Initialize autoresearch by selecting a validation mode and let the system persist state across iterations.

Frequently Asked Questions about autoresearch

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

FAQPage Schema

How do I automate persistent stateful research loops for long-running AI workflows?▼

Automating persistent stateful research loops requires a validator-gated system that stores mode and results in a dedicated .omx/state path, ensuring progress is saved across iterations for long-running AI workflows.

What is the difference between mission-script validation and prompt-architect artifacts in AI research?▼

Mission-script validation and prompt-architect artifacts are two selectable validation modes for AI research. The system chooses between them at initialization to gate artifact completion and evaluate workflow progress differently.

How do I initialize a validated research loop with selectable evaluation modes?▼

To initialize a validated research loop, select a validation mode at init-time. The system then applies artifact-based completion gating and persists the chosen mode and subsequent results in .omx/state.

Can I use artifact-based completion gating for production-grade AI experimentation?▼

Artifact-based completion gating is suitable for production-grade AI experimentation. It enforces a no-direct-CLI surface policy and uses a stateful validator to ensure safe, repeatable evaluation workflows.

Why does my automated AI research workflow need a no-direct-CLI surface policy?▼

A no-direct-CLI surface policy is needed to ensure robust, production-grade experimentation. It prevents manual evaluation surface interference by enforcing state persistence and validator-gated loops within the system.

What are the limitations of using a stateful validator-gated loop for AI research?▼

The limitation of a stateful validator-gated loop is that it requires selecting between mission-script validation and prompt-architect artifacts at initialization, restricting dynamic mode switching during the ongoing research cycle.