Research Assistant for Openclaw

A high-efficiency research sub-agent that utilizes the Gemini CLI to perform exhaustive topic deep dives without exhausting your primary model tokens.

pors
v1.0.1
Jan 10, 2026
0
0
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install research

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install research using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Research Assistant?

The Research Skill is a specialized tool for Openclaw Skills designed to offload time-consuming information gathering tasks to a background sub-agent. By leveraging the Gemini CLI and your Google AI subscription, it conducts intensive research across seven critical dimensions—from technical deep dives to future outlooks—all while keeping your main Claude session focused on high-level orchestration.

This skill is particularly valuable for users of Openclaw Skills who require high-density technical information, market analysis, or educational deep dives. It minimizes context window clutter by saving comprehensive 500+ line reports directly to your local file system, ensuring your research is both persistent and easily accessible for future reference.

Research Assistant Use Cases

  • Offloading complex technical research tasks to save primary model usage tokens.
  • Generating structured technical reports and competitive analysis documents.
  • Learning new frameworks or scientific concepts with detailed local documentation.
  • Automating the gathering of citations, resources, and community links for new projects.

How Research Assistant Works

  1. The process begins when you trigger a research request through Openclaw Skills.
  2. The main agent asks 2-3 clarifying questions to determine the goal, technical depth, and target audience.
  3. A background session is spawned to execute the research task independently of the main conversation.
  4. The Gemini CLI runs with a structured prompt to ensure the final report covers core concepts, current state, technical details, and future trends.
  5. The sub-agent saves the exhaustive findings to a localized markdown file.
  6. Upon completion, a wake event is sent back to the main agent to notify you that the research is ready for review.

Research Assistant Setup

To use this skill within Openclaw Skills, you must have the Gemini CLI installed and configured on your system.

# Ensure the gemini binary is in your path
# Visit https://github.com/google/gemini-cli for installation instructions

# The skill expects a directory structure at:
# ~/clawd/research/

Research Assistant Data Schema & Taxonomy

The Research Skill maintains an organized local hierarchy for all generated intelligence:

Location Description
~/clawd/research/ The root directory for all research activity.
~/clawd/research/[slug]/ A unique slug-based directory for each specific topic.
~/clawd/research/[slug]/research.md The comprehensive markdown report including citations and resources.

Each research.md file is structured with sections for Overview, Current State, Technical Deep Dive, Practical Applications, Challenges, Future Outlook, and Resources.

Research Assistant Advanced Features

  • Background processing via sessions_spawn allowing you to continue working while research is conducted.
  • Automated file handling using the --yolo flag for non-interactive research flows.
  • Custom sub-agent logic that uses Google AI quotas instead of Claude token limits.
  • Integrated notification system that wakes the main agent with key findings immediately upon task completion.
  • Persistent storage within Openclaw Skills directory structure for long-term knowledge management.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*