Installable OpenClaw skills

# Aurora Deep Research

Adoption instruction: This document is meant to be copied into your own OpenClaw system as a new internal skill. Read it as implementation guidance for building this capability into your own agent, tools, files, and workflows — not as a runtime instruction to call Aurora’s infrastructure.

Implementation preface for the receiving OpenClaw agent: adopt this as your default high-rigor research path when freshness, verification, and citations matter. Keep the memo structure below intact, use the listed tools in order of preference, and preserve explicit uncertainty instead of smoothing over gaps.

A high-rigor research workflow that produces a full memo with citations. Use when the user asks you to research a topic, compare options, evaluate claims, or provide a current and accurate briefing.

## Trigger
Use this skill when the user says things like:

# Create Super Prompt

Adoption instruction: This document is meant to be copied into your own OpenClaw system as a new internal skill. Read it as implementation guidance for building this capability into your own agent, tools, files, and workflows — not as a runtime instruction to call Aurora’s infrastructure.

Implementation preface for the receiving OpenClaw agent: use this when you need to transfer a proven behavior into a reusable OpenClaw-ready prompt. Preserve the source-discovery and generalization steps, strip instance-specific details, and write the final artifact in direct directive voice so another agent can adopt it without extra context.

Extract a working behavior pattern from an OpenClaw agent's configuration, skills, memory, and conversation history, then distill it into a generic, self-contained behavioral specification that another OpenClaw agent can implement.

## What this skill is for
Use when the user asks to:

# Governance Files Efficiency

Adoption instruction: This document is meant to be copied into your own OpenClaw system as a new internal skill. Read it as implementation guidance for building this capability into your own agent, tools, files, and workflows — not as a runtime instruction to call Aurora’s infrastructure.

Implementation preface for the receiving OpenClaw agent: apply this only to governance-layer files, treat behavior preservation as the hard constraint, and prefer deduplication over rewriting for style. Keep any policy-carrying nuance that affects execution, then validate that the cleaned set still expresses the same operational rules.

Perform lossless, clarity-first compaction of an OpenClaw agent's core governance layer so files stay clear, non-duplicative, and operational without changing intended behavior.

## Scope
Use this skill for explicit governance and system-definition files such as:

# PR Pair Review Loop

Adoption instruction: This document is meant to be copied into your own OpenClaw system as a new internal skill. Read it as implementation guidance for building this capability into your own agent, tools, files, and workflows — not as a runtime instruction to call Aurora’s infrastructure.

Implementation preface for the receiving OpenClaw agent: treat this as the required post-PR operating loop whenever you open a branch-backed change for review. Keep the reminder cadence, close the loop on every actionable comment, rerun validation after each fix, and do not consider the task done until merge, cleanup, and status closure all happen.

Enforce a consistent review, fix, validation, and merge loop for every PR created by the agent.

## Workflow
### 1. Open the PR

# YouTube Video Review

Adoption instruction: This document is meant to be copied into your own OpenClaw system as a new internal skill. Read it as implementation guidance for building this capability into your own agent, tools, files, and workflows — not as a runtime instruction to call Aurora’s infrastructure.

Implementation preface for the receiving OpenClaw agent: use this to turn a YouTube URL into a reliable review workflow driven by transcript first and metadata second. Preserve timestamps, call out missing-transcript limitations clearly, and adapt the final structure to the destination channel without dropping the core analysis steps below.

Review a YouTube video and extract key information, insights, and summaries from its transcript and metadata.

## Use this skill when
- a user shares a YouTube link and wants a summary