ToolTrust
C38/100
Smithery

hyperplexity-production

mcpsmithery

@Smithery

# Hyperplexity — AI Research Tables with Citations and Confidence Scores **Turn a prompt into a verified, cited research table. Or fact-check any table or document you already have.** Hyperplexity gives Claude the ability to conduct deep, structured research across entire domains — not one question at a time, but hundreds of cells simultaneously, each with a sourced answer, a confidence rating, and links to the evidence. --- ## What it does You describe what you want in plain English. Hyperplexity researches it, validates every fact, and returns a structured table where each cell has: - ✅ A verified answer - 📊 A confidence score (HIGH / MEDIUM / LOW) - 🔗 Citations to the actual sources The output is an Excel file and a shareable interactive viewer. Claude drives the full workflow — you just review and approve. --- ## Three ways to start Once installed, tell Claude what you want in plain English — it drives everything from there. | Goal | What to say to Claude | |------|----------------------| | **Generate a table from a prompt** | *"Use Hyperplexity to create a table of the top 20 US biotech companies with columns: company, market cap, lead drug, and phase."* | | **Validate an existing Excel or CSV** | *"Use Hyperplexity to validate `companies.xlsx`. Interview me about what each column means, then run the preview."* | | **Fact-check a document or passage** | *"Use Hyperplexity to fact-check this analyst report."* *(paste text or share a file path)* | Claude handles the full workflow — upload, configure, preview, approve — pausing only when a decision genuinely needs you. --- ## What each workflow does ### Generate a table from a prompt Claude starts a table-maker session, clarifies the structure if needed, builds the table, runs a free 3-row preview, and waits for your approval before billing. ### Validate an existing Excel or CSV file Upload any table and Hyperplexity fact-checks every cell against live sources. It learns the meaning of your columns through a short interview (or you can skip the interview entirely with a one-line description), then runs the same preview-then-approve flow. ### Fact-check a document or text passage Paste an analyst report, a research abstract, or any text with factual claims. Hyperplexity extracts each claim, verifies it independently, and returns support levels: **SUPPORTED / PARTIAL / UNSUPPORTED / UNVERIFIABLE** — with the source for every verdict. ### Refresh a table you already ran Re-run validation on any prior job to pick up changes in source data. No re-upload or configuration needed. --- ## Why it's different Most AI tools answer one question at a time. Hyperplexity answers a whole research domain at once — running hundreds of targeted searches, applying QC passes, and reconciling conflicting sources — then packages the results into a structured, citable format. The MCP integration means Claude can drive the entire workflow autonomously: upload, configure, preview, approve, and retrieve results — pausing only when a decision genuinely requires human input. --- ## Pricing | | Cost | |--|--| | Preview (first 3 rows) | Free | | Standard validation | ~$0.05 / cell | | Advanced validation | up to ~$0.25 / cell | | Minimum per run | $2.00 | New accounts get **$20 in free credits** — enough for several full tables. --- ## Get started 1. Get your API key at [hyperplexity.ai/account](https://hyperplexity.ai/account) 2. Install the MCP server: **Option A — uvx (recommended):** ```bash claude mcp add hyperplexity uvx mcp-server-hyperplexity \ -e HYPERPLEXITY_API_KEY=hpx_live_your_key_here ``` **Option B — Direct HTTP to Railway:** ```bash claude mcp add hyperplexity \ --transport http \ https://mcp-server-hyperplexity-production.up.railway.app/ \ --header "X-Api-Key: hpx_live_your_key_here" ``` **Option C — Smithery** (for OpenClaw and other Smithery-compatible clients): ```bash npx -y @smithery/cli@latest login npx -y @smithery/cli@latest mcp add hyperplexity/hyperplexity --client claude-code ``` Then open your client → `/mcp` → **hyperplexity → Authenticate** → enter your API key. 3. Ask your agent: *"Use Hyperplexity to generate a table of…"* Full documentation: [eliyahu.ai/api-guide](https://eliyahu.ai/api-guide)

By Smithery | 46 findings | Scanned 4/19/2026 | tooltrust-scanner/v0.3.8

12 High7 Medium9 Low18 Info

Risk Summary

Needs Approval

Excessive Permissions plus Dep Visibility raises enough risk that this tool should not be auto-trusted.

Potential impact: The agent may gain overly broad access to files, network, databases, or execution capabilities.

Recommended action: Keep this tool behind manual approval and avoid unattended runs until the risky capabilities are narrowed or removed.

Suggested policy: keep this tool behind manual approval, do not allow unattended runs, and re-scan after narrowing risky permissions.

Security Findings (46)

  • HighAS-002

    ⚠️Excessive Permissions ×11

    tool declares network permission

    upload_filecreate_jobget_job_messageswait_for_jobget_resultsget_reference_resultsreference_checkrefine_config

    tool declares exec permission

    wait_for_jobget_conversationwait_for_conversation

    Fix: Tool requests broad permissions (exec/fs/network). Validate input parameters using Enums where possible, and restrict file system operations to explicit allowed directories.

  • HighAS-003

    🔀Scope Mismatch

    get_conversation:tool name "get_conversation" implies read-only operation but declares exec permission

    Fix: Ensure tool names, descriptions, and permission declarations are internally consistent. Use explicit naming conventions that fully reflect actual capabilities.

  • MediumAS-002

    ⚠️Excessive Permissions ×7

    tool declares fs permission

    upload_fileconfirm_uploadcreate_jobwait_for_jobget_resultsupdate_tablereference_check

    Fix: Tool requests broad permissions (exec/fs/network). Validate input parameters using Enums where possible, and restrict file system operations to explicit allowed directories.

  • LowAS-011

    ℹ️Missing Rate-Limit / Timeout ×8

    tool performs network or execution operations but declares no rate-limit, timeout, or retry configuration

    upload_filecreate_jobget_job_messagesget_resultsget_reference_resultsreference_checkget_conversationrefine_config

    Fix: Declare explicit rate-limit, timeout, and retry configuration for all network and execution tools. Implement exponential back-off and surface resource state to the calling agent.

  • LowAS-002

    ⚠️Excessive Permissions

    create_job:tool declares http permission

    Fix: Tool requests broad permissions (exec/fs/network). Validate input parameters using Enums where possible, and restrict file system operations to explicit allowed directories.

  • InfoAS-014

    ℹ️Dependency Inventory Unavailable ×18

    Tool did not expose metadata.dependencies or repo_url, so supply-chain coverage is limited.

    get_balanceget_usageupload_fileconfirm_uploadcreate_jobget_job_statusget_job_messageswait_for_jobapprove_validationget_resultsget_reference_resultsupdate_tablereference_checkstart_table_makerget_conversationsend_conversation_replyrefine_configwait_for_conversation

    Fix: Review and remediate the identified issue.

Scan this tool yourself

Reproduce this audit locally, integrate into CI, or let your agent audit its own tools.

Install once, then scan any MCP server:

$ curl -sfL https://raw.githubusercontent.com/AgentSafe-AI/tooltrust-scanner/main/install.sh | bash
$ tooltrust-scanner scan --server "npx -y hyperplexity-production"

Adjust the package name if your npm registry name differs from the tool ID. View source

Add badge to your README

Copy this Markdown to show your ToolTrust grade on GitHub.

[![ToolTrust Grade C](https://raw.githubusercontent.com/AgentSafe-AI/tooltrust-directory/main/docs/badges/grade-c.svg)](https://github.com/AgentSafe-AI/tooltrust-directory)