ok-computer-swarm

# OK Computer Swarm

## Overview

This skill lets OpenClaw emulate the “100 sub‑agents” style of Kimi’s Agent Swarm【453334500861599†L40-L99】.  When you need to research several topics at once, the skill
spins up lightweight sub‑agents that fetch the top web results via DuckDuckGo.  By running these tasks in parallel, the skill reduces overall waiting time and surfaces a diverse set of sources.

## Commands

### `/ok-computer-swarm search`

Run concurrent searches for multiple topics.

**Inputs**

- `query` (string, repeated):  One or more search phrases.  You can provide multiple `query` flags to search many topics at once.  At least one `query` is required.

**Example**

```bash
python scripts/swarm_search.py --query "Agent Swarm" --query "OpenClaw skills"
```

**Output**

The script prints a JSON array where each element corresponds to a search query.  Each element contains the original query and an array of result objects (title and URL).  The format is easy for downstream agents to parse and can be further processed or summarised.

## When to use this skill

Use `ok-computer-swarm` whenever you need to gather high‑level information on multiple topics concurrently.  It is ideal for:

- Broad research tasks that involve several different subjects.
- Generating a starting point for more in‑depth analysis.
- Situations where time is critical and sequential research would be too slow.

## Limitations

- The skill uses DuckDuckGo’s free API; results may be less comprehensive than paid search APIs.
- It performs minimal summarisation.  Consider integrating additional summarisation or reading tools if you need deeper insights.

---

# OK Computer Swarm Skill

![Project Logo](./3000logo.png)

This repository contains a **ClawHub skill** that emulates Kimi.com’s *OK Computer* agent experience using community‑accessible components.  The goal of the skill is to demonstrate how a single command can orchestrate multiple sub‑agents to perform research in parallel and then synthesize the results into a cohesive report.  This style mirrors Kimi’s Agent Swarm, which spawns dozens of sub‑agents that each explore a different facet of a problem before regrouping to reconcile their findings【453334500861599†L40-L99】.

## What the skill does

When invoked, the skill accepts a list of topics or questions.  For each topic, it creates a *sub‑agent* that uses the DuckDuckGo API to fetch high‑level information.  These sub‑agents run concurrently, dramatically reducing the time required to gather a broad set of facts.  Once all sub‑agents complete their searches, the primary agent collates the data and returns a structured JSON report.  The report includes the original query, a list of the top results (title and URL) and a short synopsis for each topic.

> **Note:**  This skill is meant as a starting point.  You are encouraged to expand the sub‑agents with richer tooling—such as reading full articles, performing summarisation, or calling additional skills.  The concurrent architecture makes it straightforward to add more complex behaviours later.

## Repository contents

- `SKILL.md` – Metadata and usage instructions for OpenClaw.  This file defines the name, description and examples so that OpenClaw knows when to call the skill.
- `_meta.json` – Version and ownership metadata used by ClawHub.  Edit the `owner` and `slug` fields before publishing.
- `scripts/swarm_search.py` – The Python implementation of the skill.  It spawns worker threads to search DuckDuckGo and returns a JSON report.  Run `python scripts/swarm_search.py --help` for usage.
- `requirements.txt` – Python dependencies needed to run the skill.
- `...