skills.rest
solitude6060
solitude6060COMMUNITY
Explore all skills by this author →
WebsiteGitHub

autoresearch

Iterate a single mission through evaluator-gated experiments until terminal conditions.

Updated May 18, 2026
Search Tags:#automation#iterative improvement#stateful workflows#incident triage#evaluator contract#mission-driven#experiment logs
One-click install
npx skills add https://github.com/solitude6060/Yao-skills --skill autoresearch-solitude6060
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: autoresearch
Source: https://github.com/solitude6060/Yao-skills/tree/main/skills/autoresearch
Command: npx skills add https://github.com/solitude6060/Yao-skills --skill autoresearch-solitude6060

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It solves the problem of getting stuck on one-pass answers by running a bounded, evaluator-driven loop that keeps improving a single mission until time limits or explicit stop conditions are hit.

Core Features & Use Cases

  • Stateful single-mission improvement loop: Keeps ownership of exactly one mission at a time and iterates through non-passing results.
  • Strict evaluator contract: Requires evaluator output as structured JSON with a boolean pass (and optional numeric score) so decisions stay consistent.
  • Durable decision logs and artifacts: Persists per-iteration evaluation JSON and human-readable markdown logs under .omc/autoresearch/ and/or .omc/logs/autoresearch/.
  • Cron-friendly reruns: Supports periodic scheduled mission enhancement by appending new run artifacts rather than overwriting history.

Quick Start

Run deep-interview with autoresearch to set up the mission and evaluator, then let the skill iterate and persist logs under .omc/autoresearch/ until the max-runtime ceiling is reached.

Frequently Asked Questions about autoresearch

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

FAQPage Schema
How do I automate iterative improvement for a single coding mission?▼

Stateful iterative improvement works by running bounded, evaluator-gated experiments that persist per-iteration artifacts and markdown logs, stopping only when the evaluator returns a passing JSON result or reaches max-runtime.

What's the best way to set up an evaluator contract for automated code review refinement?▼

Set up an evaluator contract by defining a structured JSON schema requiring a boolean `pass` field and an optional numeric `score`, ensuring consistent gating decisions across all iterative improvement loops.

Can I schedule periodic incident triage follow-ups without overwriting previous run logs?▼

Periodic incident triage follow-ups are cron-friendly, appending new run artifacts and durable decision logs under `.omc/autoresearch/` rather than overwriting previous history.

Does stateful workflow automation support bounded project-health checks?▼

Stateful workflow automation supports bounded project-health checks by enforcing a strict stop policy, converging iterative evaluations under max-runtime limits while persisting per-iteration artifacts.

Why does my mission improvement loop stop before the evaluator passes?▼

A mission improvement loop stops before passing because it enforces strict terminal conditions such as max-runtime, halting iterative experiments even if the evaluator JSON has not yet returned a passing result.

Related Skills

Explore More Software Engineering Skills→
xz1220xz1220

autoresearch

Run iterative improvement loops with JSON evaluation and decision logs.

Community
Advanced
jimmi2051jimmi2051

autoresearch

Run evaluator-driven mission improvement loops with JSON logs and Markdown decision records.

Community
Advanced
RobinNorbergRobinNorberg
5

autoresearch

Run bounded iterative improvement loops with evaluator-driven logs and cron reruns.

Community
Advanced
Leap0920Leap0920
1

autoresearch

Iteratively improves a task until a bounded stop condition using JSON evaluator feedback.

Community
Advanced