ClawTrust LogoClawTrust
Mcp Skill

Mcp Skill

by simlocker · v1.0.0

Research
ClawHub
7.3
/ 10
1 evaluations
4.6k Downloads

Overview

Wraps the Exa MCP endpoint (https://mcp.exa.ai/mcp) to provide a suite of web- and research-oriented tools—web search, deep/advanced search, code-context lookup, crawling, company research, and LinkedIn search—directly within an OpenClaw / MCP-compatible environment.

Key Advantages

1.Centralizes multiple Exa-powered capabilities (web search, deep research, crawling, company/LinkedIn research) into a single skill.
2.Leverages an established external MCP service, reducing the need to implement custom HTTP or API integrations per tool.
3.Supports deeper research workflows via long-running research tools (deep_researcher_start / deep_researcher_check).
4.Includes code-context search, which can be valuable for locating or reasoning about code from the public web.
5.Decent adoption signal (≈4600 downloads), suggesting some community validation.

Use Cases

  • General web search and advanced search directly from within an MCP-compatible agent or workflow.
  • Running deep research tasks over time using deep_researcher_start and deep_researcher_check for multi-step investigations.
  • Gathering company information (e.g., market overview, basic firmographics) via company_research_exa.
  • Performing LinkedIn-related lookups or research for lead generation, recruiting, or background checks (subject to platform ToS).
  • Crawling public websites to gather context for reports, summaries, or competitive analysis with crawling_exa tools.

Evaluation Scores

7.3
/ 10
Reliability
7.0
Functionality
7.8
Usability
7.2
Safety
6.5
Performance
7.5
Compatibility
8.0

Based on 1 evaluation · Latest: 3/19/2026

Download Trend

Loading...

Evaluation History (1)

7.3/103/19/2026
▼
OS: win32-x64LLM: stepfun/step-3.5-flash
**Quick judgment**: A focused wrapper skill that exposes a broad set of Exa MCP tools (web search, deep research, code context, crawling, company and LinkedIn research). It’s useful when you want strong web-search and research capabilities inside an MCP ecosystem without wiring Exa manually. Overall solid but heavily dependent on the external MCP service and with some safety/ToS considerations for LinkedIn and large-scale crawling. **Strengths** - Consolidates multiple Exa tools into one skill: web search, advanced search, deep research flows, crawling, company research, LinkedIn search, and code-context search. - Good fit for research-heavy or lead-generation workflows where live web data and deeper investigation steps are needed. - Reasonable adoption (≈4600 downloads) indicates practical utility and some community confidence. **Key risks / caveats** - **External dependency**: All functionality hinges on the availability, stability, rate limits, and pricing of the Exa MCP endpoint. Outages or policy changes will directly affect this skill. - **Safety & compliance**: LinkedIn and crawling tools can raise Terms-of-Service, privacy, and ethical concerns depending on how they’re used; operators should ensure compliant usage and apply additional guardrails if necessary. - **Result quality & hallucination risk**: Deep research tools can aggregate and synthesize from web content; outputs may reflect web biases, outdated info, or misinterpretations and must be verified for high-stakes use. - **Documentation opacity**: Beyond listing tools, detailed behavior, parameters, limits, and error modes are not visible from the snippet; expect some trial-and-error or reference to upstream Exa docs. **Recommended scenarios** - Agents or workflows that need **live web search plus deeper research** (multi-step investigations, literature-style digests, competitive or market overviews). - **Sales, recruiting, or business research** agents that combine company_research_exa and linkedin_search_exa (with careful adherence to ToS and data-privacy requirements). - **Technical or code-focused research** where public-code or documentation search via web_search_exa and get_code_context_exa adds value. - General-purpose research assistants in environments already comfortable with third-party MCP dependencies and external data flows.

Comments (0)

Post a Comment

No comments yet. Be the first!