← Back to Skills Marketplace
139
Downloads
1
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install openwebsearch
Description
Single entry skill for open-websearch setup and focused live retrieval, preferring local CLI/daemon paths while remaining compatible with workspace-exposed M...
Usage Guidance
This skill is an instruction-only helper and appears coherent with its stated purpose. It does not demand credentials or install code by itself. Before you proceed, be aware it may ask you to: (a) run local commands like `open-websearch serve`/`status`, (b) allow package installs (e.g., Playwright via npm) or start a local daemon, and (c) provide proxy or endpoint details. Only approve installs or long-running daemon startups if you trust the local package sources and understand network/proxy implications. If you are uncomfortable granting those actions, deny or review each confirmation the skill requests. If you want extra assurance, verify available local binaries and any repository checkouts yourself rather than letting the agent perform installs.
Capability Analysis
Type: OpenClaw Skill
Name: openwebsearch
Version: 1.0.1
The openwebsearch skill bundle provides a legitimate framework for an AI agent to perform web searches and fetch content via local CLI tools or MCP. The instructions in SKILL.md and references/setup.md include explicit safety rules to protect the agent from prompt injection in external web content and require user confirmation before performing high-impact actions like package installation (npm) or daemon startup. No evidence of malicious intent, data exfiltration, or unauthorized execution was found.
Capability Assessment
Purpose & Capability
The name/description (open-websearch setup and focused retrieval) aligns with the instructions: detect local CLI/daemon or MCP-exposed tools, prefer the smallest path, validate availability, and perform focused searches/fetches. The allowed tools (search, fetchWebContent, fetchGithubReadme) are appropriate for the stated purpose.
Instruction Scope
SKILL.md contains concrete runtime actions (checking for a local command, running explicit daemon commands like `open-websearch serve` and `open-websearch status`, and guidance for Playwright/browser installation). These are within scope for configuring/using a local websearch agent, and the skill explicitly requires user confirmation before package installs, daemon startup, or config edits. The guidance references environment variables (e.g., PLAYWRIGHT_* and FETCH_WEB_INSECURE_TLS) as configuration knobs but does not instruct blind reading/exfiltration of unrelated system files or secrets.
Install Mechanism
There is no install spec and no code files — this is instruction-only. The document only suggests possible npm installs (e.g., Playwright) but requires confirmation before doing so; no downloads or arbitrary URLs are embedded in the skill itself.
Credentials
The skill declares no required environment variables, credentials, or config paths. It references optional runtime/config env variables as configuration points (for Playwright or TLS fallback), which is proportionate to the stated functionality and only used as guidance rather than mandatory secrets access.
Persistence & Privilege
always:false and default autonomy settings are used. The skill does not request permanent or elevated platform privileges, does not modify other skills' configs, and instructs the agent to prompt before starting daemons or changing workspace/client config.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install openwebsearch - After installation, invoke the skill by name or use
/openwebsearch - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
No user-visible changes in this version.
- No changes detected between versions 1.0.1 and 1.4.0; the skill logic and documentation remain unchanged.
- Version number increment only; behavior and features are the same as the previous version.
v1.0.0
open-websearch 1.4.0
- Now prefers local CLI/daemon onboarding and retrieval when available, while preserving compatibility with MCP-based setup and activation.
- Clarifies setup and activation workflow; distinguishes between missing, inactive, and active states.
- Updates research and retrieval guidance: prioritizes direct URL fetching, focused search, and minimal necessary actions.
- Reinforces safety and reliability rules for interacting with external web content.
- Expands engine selection recommendations and fallback behaviors for search modes and error handling.
- Improves messaging when open-websearch capability is missing or requires setup.
Metadata
Frequently Asked Questions
What is OpenWebSearch?
Single entry skill for open-websearch setup and focused live retrieval, preferring local CLI/daemon paths while remaining compatible with workspace-exposed M... It is an AI Agent Skill for Claude Code / OpenClaw, with 139 downloads so far.
How do I install OpenWebSearch?
Run "/install openwebsearch" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is OpenWebSearch free?
Yes, OpenWebSearch is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does OpenWebSearch support?
OpenWebSearch is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created OpenWebSearch?
It is built and maintained by Aasee (@aas-ee); the current version is v1.0.1.
More Skills