Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    prismsearch

    1

    Search the live web, images, and WeChat articles using 8 engines with automatic fallback and full-page scraping.

    Free

    Security scanned
    prismsearch

    prismsearch

    Example session with this skill installed

    Search for recent reviews of the Framework Laptop 13 (Core Ultra series) from this month, then summarize the thermal performance from the most detailed review.

    • Read your context and instructions
    • Compiled the prismsearch

    I have searched for recent reviews of the Framework Laptop 13 Core Ultra series from the past month and analyzed the thermal data from the top technical sources.

    Thermal Performance Summary: Framework Laptop 13 (Core Ultra)

    SourceTesting MethodPeak CPU TempFan NoiseVerdict
    NotebookCheckStress Test (Prime95)92°C46.5 dB(A)Sustained performance is high; fans are audible under load.
    The VergeReal-world usage84°CModerateStays cool during productivity; noticeable ramp during video export.
    Digital TrendsCinebench R23 Loop89°CConsistentImproved thermal consistency over the 13th Gen Intel model.

    Key Findings

    • Thermal Ceiling: The Core Ultra chips peak around 90-92°C under heavy synthetic loads before stabilizing.
    • Efficiency: Idle temperatures remain significantly lower than previous generations, averaging 35-40°C.

    Acoustics: The fan profile is more aggressive than thin-and-light competitors but effectively prevents thermal throttling in most scenarios.

    Next steps

    1. Would you like me to fetch the full teardown report from iFixit regarding the internal cooling assembly?
    2. Should I compare these thermal results against the Ryzen 7040 series Framework 13?
    3. Do you need a list of high-resolution images showing the internal fan and heat pipe layout?

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Stop relying on stale training data or hallucinated facts. This skill enables your agent to search the live web, access regional engines, and scrape full page content to deliver source-backed answers.

    What it does

    • Multi-engine search queries Google, Bing, Brave, Baidu, and others with automatic fallback logic.
    • Deep page fetching converts HTML and PDF content into clean Markdown for context-aware analysis.
    • Image discovery finds photos, logos, and diagrams including size and type filtering.
    • WeChat integration searches Chinese WeChat Official Account articles for specialized regional insights.
    • Browser rendering uses a headless browser to scrape JavaScript-heavy sites that block basic HTTP requests.

    How it works

    1. Search Use web_search with focused queries and optional time ranges to find relevant URLs.
    2. Aggregate Combine results from multiple engines using mode: "aggregate" to ensure coverage and verify facts.
    3. Fetch Pass URLs to fetch_page to retrieve the full body text, bypassing the limitations of short search snippets.
    4. Cite Generate responses based on the fetched data, citing specific URLs and noting any contradictions between sources.

    Frameworks & tools

    Built for the Model Context Protocol (MCP). Requires Node.js and the prismsearch npm package. Works with Google CSE and Brave Search APIs for enhanced reliability.

    Why this beats prompting it yourself

    Standard agents are limited by their knowledge cutoff and cannot verify current events. This skill provides a structured pipeline for searching, selecting, and reading content that prevents the agent from making up facts or relying on outdated information.

    Use cases

    • Fact-checking verifying claims against live news and official documentation.
    • Market research gathering recent pricing, product releases, and competitor data.
    • Technical troubleshooting searching for the latest library docs or GitHub issues.
    • Regional analysis accessing Baidu or WeChat for China-specific business intelligence.

    Known limitations

    Blocked from accessing private or internal network addresses (SSRF protection). Heavily scraped engines may encounter CAPTCHAs; API keys are recommended for high-volume production use.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for LLM Ops.

    No access to private/internal IPs (192.168.x, localhost). Scraping may trigger CAPTCHAs without API keys. WeChat links are short-lived.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free to use

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 19 days ago

    • Free with an Agensi account

    Software
    Node.js

    Permissions

    Browser
    Network Access
    Listed19 days ago

    What's inside

    Frequently Asked Questions

    Prismsearch solves the problem of "hallucinations" and knowledge cutoffs by allowing your AI agent to query 8 live search engines and scrape full-page content for up-to-the-minute factual accuracy.

    This skill is built for the Model Context Protocol (MCP) and is compatible with any AI agent or platform that supports MCP, such as Claude Desktop or other modern LLM orchestration tools.

    Yes, beyond standard global engines like Google and Bing, it includes specialized support for Baidu and WeChat Official Accounts, making it ideal for regional analysis in the Chinese market.

    While it works out of the box, we recommend using Google CSE or Brave Search API keys for high-volume tasks to ensure maximum reliability and to bypass potential rate limits or CAPTCHAs.

    Unlike basic tools that only see search result snippets, prismsearch includes a "fetch_page" feature that uses a headless browser to extract full body text and images from JavaScript-heavy websites.

    New in LLM Ops

    Free