@cite42/mcp 0.1.3 → 0.1.4

package/README.md

@@ -1,52 +1,31 @@
 # @cite42/mcp
 
-MCP (Model Context Protocol) server for the [Cite42](https://cite42.dev) API, built on
-[Mastra](https://mastra.ai).
+Stdio MCP server for [Cite42](https://cite42.dev).
 
-Cite42 lets Claude Desktop, Cursor, Claude Code, VS Code, Windsurf, and other MCP clients query
-AI-search visibility, citations, SEO keyword data, Google Trends, Reddit/YouTube signals, and
-creator/agency workflow contracts.
+Use Cite42 from your coding agent to:
 
-This package is a thin MCP client. It forwards tool calls to the hosted Cite42 REST API, so billing,
-metering, provider keys, caching, and rate limits stay server-side.
+- check AI search answers across ChatGPT, Claude, Perplexity, Gemini, and Google AI Overview
+- find citations, rankings, brand comparisons, and sentiment
+- get SEO keyword data, Google Trends, Reddit trends, and YouTube trends
+- run Cite42 workflows for content, demand, prompts, competitors, and citation gaps
 
-## Requirements
+Requires a Cite42 API key from [cite42.dev/app/keys](https://cite42.dev/app/keys).
 
-- Node.js 18 or newer
-- A Cite42 API key
-- Prepaid Cite42 credits
+## Claude Code
 
-Get an API key from [cite42.dev/app/keys](https://cite42.dev/app/keys). Top up credits at
-[cite42.dev/app/billing](https://cite42.dev/app/billing).
-
-Without an API key, the MCP server starts but tool calls cannot reach the Cite42 API.
-
-## Install
-
-You do not need to install this package globally. Use `npx`; your MCP client will fetch and run the
-latest published version.
-
-## Configure Claude Desktop
-
-Open your Claude Desktop config:
-
-- macOS: `~/Library/Application Support/Claude/claude_desktop_config.json`
-- Windows: `%APPDATA%\Claude\claude_desktop_config.json`
+```bash
+claude mcp add cite42 -e CITE42_API_KEY=cite42_live_your_key_here -- npx -y @cite42/mcp
+```
 
-Add the Cite42 MCP server:
+## Codex CLI
 
-```json
-{
-  "mcpServers": {
-    "cite42": {
-      "command": "npx",
-      "args": ["-y", "@cite42/mcp", "--api-key=cite42_live_your_key_here"]
-    }
-  }
-}
+```bash
+codex mcp add cite42 --env CITE42_API_KEY=cite42_live_your_key_here -- npx -y @cite42/mcp
 ```
 
-You can also pass the key as an environment variable:
+## Cursor
+
+Add this to your MCP config:
 
 ```json
 {
@@ -54,103 +33,35 @@ You can also pass the key as an environment variable:
     "cite42": {
       "command": "npx",
       "args": ["-y", "@cite42/mcp"],
-      "env": { "CITE42_API_KEY": "cite42_live_your_key_here" }
+      "env": {
+        "CITE42_API_KEY": "cite42_live_your_key_here"
+      }
     }
   }
 }
 ```
 
-Restart your MCP client after editing the config.
-
-## Configuration
-
-| Setting | Flag | Env var | Default |
-| --- | --- | --- | --- |
-| API key | `--api-key=cite42_live_...` | `CITE42_API_KEY` | Required |
-| API base URL | `--api-base=...` | `CITE42_API_BASE` | `https://api.cite42.dev/v1` |
-
-Flags win over environment variables when both are set.
-
 ## Tools
 
-| Tool | What it does |
-| --- | --- |
-| `cite42_search` | Runs a query across ChatGPT, Claude, Perplexity, Gemini, and Google AI Overview, returning answers and cited sources. |
-| `cite42_citations` | Aggregates cited URLs for a query and can check whether a specific URL is cited. |
-| `cite42_rankings` | Measures competitor/brand mention rate, average position, and per-model ranking breakdown. |
-| `cite42_compare` | Compares one brand against competitors, including co-mention data. |
-| `cite42_sentiment` | Scores positive/neutral/negative brand sentiment with supporting phrases. |
-| `cite42_keywords` | Returns search volume, CPC, competition, competition level, and optional curated keyword ideas. |
-| `cite42_trends` | Returns Google Trends interest, related/rising queries, and trend direction. |
-| `cite42_reddit_trends` | Returns Reddit audience questions, pain points, threads, phrases, mentioned products, and URLs. |
-| `cite42_youtube_trends` | Returns YouTube rising videos, creator angles, title patterns, channels, and opportunities. |
-| `cite42_find_content_opportunities` | Combines AI search, SEO keywords, trends, Reddit, YouTube, and citations. |
-| `cite42_analyze_topic_demand` | Combines SEO demand, trends, AI answers, and audience conversations for a topic. |
-| `cite42_map_ai_prompts` | Discovers buyer prompts worth tracking from AI, SEO, trend, Reddit, and YouTube data. |
-| `cite42_analyze_competitor_content` | Combines AI rankings, comparisons, citations, keyword demand, and topic coverage. |
-| `cite42_find_ai_citation_gaps` | Finds AI queries where competitors are cited but your brand/domain/URL is not. |
-| `cite42_generate_content_brief` | Builds brief-ready data from AI answers, citations, SEO demand, trends, Reddit, and YouTube. |
-
-## Model Selection
-
-The AI-search tools accept an optional `models` array:
-
-```json
-{
-  "query": "best customer support software for B2B SaaS",
-  "models": ["perplexity", "gemini"]
-}
-```
-
-Omit `models` to query all five surfaces: `chatgpt`, `claude`, `perplexity`, `gemini`, and
-`google_ai_overview`.
-
-## Keyword Ideas
-
-`cite42_keywords` can return curated keyword ideas when `ideas` is true:
-
-```json
-{
-  "seed": "customer service software",
-  "country": "us",
-  "ideas": true
-}
-```
-
-Each idea includes provider-backed metrics:
-
-```json
-{
-  "keyword": "customer support software",
-  "volume": 1900,
-  "cpc": 116.7,
-  "competition": 0.15,
-  "competitionLevel": "LOW"
-}
-```
-
-The curation step selects from provider candidates only; it does not invent new keyword strings.
-
-## Pricing
-
-Each successful tool call consumes credits from your Cite42 balance.
-
-| Tool group | Price |
-| --- | --- |
-| Search, citations, rankings, compare | Google AI Overview $0.01, Gemini $0.02, ChatGPT/Claude $0.04, Perplexity $0.05 per selected model |
-| Sentiment | Google AI Overview $0.03, Gemini $0.04, ChatGPT/Claude $0.06, Perplexity $0.07 per selected model |
-| Keywords | $0.15 per call |
-| Google Trends | $0.01 per call |
-| Reddit trends | $0.05 per call |
-| YouTube trends | $0.05 per call |
-
-Workflow tools are priced from the primitive tools they run. Check the Cite42 dashboard for current
-usage and balance.
-
-## Usage Guidance
-
-Use Cite42 for AI-search, citation, SEO, keyword, trend, and market-research data. It is not a
-general coding or writing assistant. Code-related research is fine when the goal is to compare what
-AI models recommend or cite.
-
-If a task may require many calls or may cost more than a few dollars, ask the user before continuing.
+- `cite42_search`
+- `cite42_citations`
+- `cite42_rankings`
+- `cite42_compare`
+- `cite42_sentiment`
+- `cite42_keywords`
+- `cite42_trends`
+- `cite42_reddit_trends`
+- `cite42_youtube_trends`
+- `cite42_find_content_opportunities`
+- `cite42_analyze_topic_demand`
+- `cite42_map_ai_prompts`
+- `cite42_analyze_competitor_content`
+- `cite42_find_ai_citation_gaps`
+- `cite42_generate_content_brief`
+
+## Notes
+
+- This package uses stdio MCP.
+- Tool calls use your Cite42 credits.
+- No API key means no successful tool calls.
+- Optional base URL: `CITE42_API_BASE` or `--api-base=https://...`

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@cite42/mcp",
-  "version": "0.1.3",
-  "description": "MCP server for the Cite42 API. Requires a Cite42 API key for AI search, citations, SEO keywords, trends, and workflows.",
+  "version": "0.1.4",
+  "description": "Stdio MCP server for Cite42 AI search, citations, SEO keywords, trends, and workflows.",
   "license": "MIT",
   "type": "module",
   "bin": {