Exa Web Search (Free)

PassAudited by ClawScan on May 1, 2026.

Overview

This is a coherent instruction-only Exa MCP search skill, with the main caveat that queries and optional advanced research/crawling tools use a remote Exa service.

This skill appears safe for ordinary web, code, and company research. Before installing, understand that it adds a remote Exa MCP endpoint through mcporter and that anything you ask it to search for may be sent to Exa. Be more cautious with the optional full tool set, especially crawling, people search, and deep researcher tasks.

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

Search queries, URLs, company names, or people-search terms entered through this skill may be shared with Exa.

Why it was flagged

The skill directs users to connect mcporter to Exa's remote MCP endpoint, so tool queries are sent to an external provider.

Skill content
mcporter config add exa https://mcp.exa.ai/mcp
Recommendation

Use it for public or non-sensitive research queries, and avoid sending private business, personal, or confidential data unless you are comfortable with Exa handling it.

What this means

Search results or code examples could influence the agent's recommendations even if the underlying web content is unreliable.

Why it was flagged

The skill retrieves external web and code content that may be incorporated into the agent's answer; such content can be inaccurate, outdated, or include unsafe instructions.

Skill content
Find code examples and docs from GitHub, Stack Overflow.
Recommendation

Treat retrieved web results and code snippets as untrusted references; verify sources and review code before using it.

What this means

If enabled and invoked, a remote research task may run outside the immediate chat turn and return later via a task ID.

Why it was flagged

The optional advanced configuration exposes a provider-side research-agent workflow with start and check operations, which is broader than a single search call but is clearly disclosed as optional.

Skill content
deep_researcher_start/check - AI research agent
Recommendation

Enable the advanced tool set only when needed, keep track of started research tasks, and avoid using it for sensitive topics unless acceptable under the provider's terms.