Jerry Sniffs MCP

Official MCP server for JerrySniffs that provides web search, Twitter/X search, tweet lookup, Reddit search, and URL-to-Markdown conversion tools to your LLM agents, with Instagram and TikTok searches planned for coming weeks.

Pricing

• No subscription. Simply buy credits whenever you need. Credits don't expire.
• Every 10$ gives you following:
  • Google Searches: 15,000
  • Twitter Searches: 1,500
  • Tweet Lookups: 1,500
  • Reddit Searches: 2000
  • URL to Markdown conversion: 15,000
• You can buy as many 10$ credit packs as you need. Credits add-up and do not expire.

Please refer to official Website for latest details

Installation

npm install -g jerrysniffs-mcp

MCP Client Integration

Get your API key first from dashboard (known as API Token on dashboard) and integrate as below. The MCP server runs on stdio transport.

Claude code and compatible harnesses like pi, cline, etc. (e.g. claude_desktop_config.json or mcp.json):

{
  "mcpServers": {
    "jerrysniffs": {
      "command": "npx",
      "args": ["-y", "jerrysniffs-mcp"],
      "env": {
        "API_KEY": "your-api-key-here"
      }
    }
  }
}

Note: If you are using PI agent, you might need to tell your LLM to first connect to jerrysniffs mcp by specifically asking for it. This is an issue with PI. This has to be done just the first time.

For codex, you can add to your ~/.codex/config.toml:

Note: both blocks mentioned below should be added

[mcp_servers.jerrysniffs]
command = "npx"
args = ["-y", "jerrysniffs-mcp"]
tool_timeout_sec = 300

[mcp_servers.jerrysniffs.env]
API_KEY = "your-api-key-here"

For OpenCode, you can add to your ~/.config/opencode/opencode.json:

{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "jerrysniffs": {
      "type": "local",
      "command": ["npx", "-y", "jerrysniffs-mcp"],
      "enabled": true,
      "environment": {
        "API_KEY": "your-api-key-here"
      }
    }
  }
}

Tools Available to Your Agent

Detailed docs available here: JerrySniffs API Docs

search

Search the web (google search). Returns URLs and metadata for matching results.

| Parameter | Type | Required | Description | |-----------|--------|----------|-------------| | query | string | yes | Search query string (supports operators like site:, inurl:, etc.) | | limit | number | no | Maximum number of results to return | | country | string | no | Two-letter country code (e.g. "US", "GB") | | page | number | no | Page number for paginated results | | start | number | no | Zero-based offset of the first result | | tbs | string | no | Time-based search filter (e.g. "qdr:w" for past week) |

twitter_search

Search Twitter/X for tweets matching a query. Returns matching tweets with metadata.

| Parameter | Type | Required | Description | |---------------|--------|----------|-------------| | query | string | yes | Search query for Twitter/X | | search_type | string | no | Type of search (e.g. "Latest", "Top") | | limit | number | no | Maximum number of tweets to return |

tweet_lookup

Look up a single tweet by its ID. Returns the tweet with full metadata.

| Parameter | Type | Required | Description | |-----------|--------|----------|-------------| | id | string | yes | The ID of the tweet to look up |

reddit_search_posts

Search Reddit for posts matching a query. Optionally filter by subreddit, sort order, and time range.

| Parameter | Type | Required | Description | |-------------|--------|----------|-------------| | query | string | yes | Search query for Reddit posts | | subreddit | string | no | Limit search to a specific subreddit (e.g. "SaaS") | | sort | string | no | Sort order (e.g. "top", "new", "hot", "relevance") | | time | string | no | Time range filter (e.g. "hour", "day", "week", "month", "year", "all") | | limit | number | no | Maximum number of posts to return |

url_to_markdown

Fetch a URL and convert its content to Markdown. Supports HTML pages, PDFs, DOC, and DOCX files.

| Parameter | Type | Required | Description | |-----------|--------|----------|-------------| | url | string | yes | The public URL to fetch and convert to Markdown. Must use http or https | | proxy | string | no | Optional proxy URL (e.g. "http://username:[email protected]:8080") to route the request through |

License

ISC