web-claude
ReviewAudited by ClawScan on May 10, 2026.
Overview
The skill is mostly a coherent web-search helper, but it can automatically use a logged-in Claude.ai browser session and persist search results, so it deserves review before installation.
Use this skill only if you are comfortable with searches being sent to Brave, DuckDuckGo, and potentially your logged-in Claude.ai account. Consider making the Claude.ai browser fallback opt-in, using a separate browser profile, avoiding sensitive queries, and reviewing or deleting the memory/research/ cache periodically.
Findings (4)
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.
A routine search request could automatically send the query through a logged-in Claude.ai browser session, consuming quota and creating account activity.
The skill directs automatic fallback into browser automation when other search methods fail, without stating that the user should approve that higher-impact path first.
if both failed:
fallback to web-claude browser methodRequire explicit user confirmation before using the Claude.ai browser fallback, or make the browser method opt-in only.
The agent may act through the user's Claude.ai session, consuming daily message limits and leaving searches in the user's Claude account history.
The skill relies on an authenticated Claude.ai browser profile, but the registry metadata declares no primary credential or required config path.
- claude.ai logged in (openclaw profile)
Declare the Claude.ai session requirement clearly, use a dedicated low-privilege browser profile if possible, and ask before each authenticated browser use.
Searches that include sensitive topics could remain in local memory and may be reused or read during later tasks.
The skill persistently stores queries, timestamps, links, summaries, and extracted insights for later reuse.
All search results automatically saved to `memory/research/` folder
Avoid searching secrets or private information, and periodically review or delete files under memory/research/ if they are no longer needed.
The skill may fail or require manual package setup, and users must ensure any installed package comes from a trusted source.
The instructions reference an external Python package, while the provided install metadata contains no install spec or dependency declaration.
from duckduckgo_search import DDGS
Document the required package and installation method, preferably with pinned versions and clear provenance.