Performs web searches using DuckDuckGo to retrieve real-time information from the internet. Use when the user needs to search for current events, documentation, tutorials, or any information that requires web search capabilities.

PassAudited by ClawScan on May 1, 2026.

Overview

The skill appears purpose-aligned for DuckDuckGo web search, with expected but noteworthy local command, package-install, and external search-query behavior.

This looks like a normal instruction-only DuckDuckGo search skill. Before installing, be comfortable with the agent installing an external Python package and running Python search commands, and avoid using the skill for secrets or sensitive private queries.

Findings (3)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

The agent may run local Python and package-manager commands as part of using the search skill.

Why it was flagged

The skill grants the agent local command execution through Python and package-management tools. This is expected for the documented search workflow, but it is broader than a dedicated search API call.

Skill content
allowed-tools: Bash(duckduckgo-search:*), Bash(python:*), Bash(pip:*), Bash(uv:*)
Recommendation

Install only if you are comfortable with the agent running these commands, and prefer reviewing install or command execution prompts when available.

What this means

Future or compromised package versions could change behavior compared with what the skill documentation describes.

Why it was flagged

The documented setup installs an external package without pinning a version or providing a lockfile. This is common for simple skills, but users are relying on the current package distribution.

Skill content
uv pip install duckduckgo-search

# 或使用 pip 安装
pip install duckduckgo-search
Recommendation

Consider pinning a known-good package version or reviewing the package source before installation in sensitive environments.

What this means

Search queries may leave the local environment, so private or sensitive terms could be exposed to the search provider.

Why it was flagged

The examples send search queries through the DuckDuckGo search library to an external provider. This is the core purpose of the skill, but it creates an external data flow.

Skill content
with DDGS() as ddgs:
    results = list(ddgs.text('Python tutorial', max_results=5))
Recommendation

Avoid putting secrets, confidential business details, or sensitive personal information into search queries.