A Claude Code plugin with AI research agent skills powered by MCP.
No API keys, no subscriptions, no configuration. Install and start researching.
Powered by RivalSearchMCP — 10 MCP research tools across web, social, news, GitHub, academic, and document sources.
This Claude Code plugin gives Claude a complete AI research platform — 10 MCP tools, 5 agent skills, and 6 specialist agents that work together to handle any research task.
- 10 MCP tools via RivalSearchMCP — web search, social media, news, GitHub, academic papers, document analysis, and more
- 5 agent skills — slash commands that run multi-step research workflows and produce structured reports
- 6 AI research agents — specialist analysts that Claude invokes automatically based on your task
- Portable hooks — audit logging and quality gates that travel with the plugin
Once this Claude Code plugin is installed, try these agent skills:
"Research the current state of AI code generation tools"
/rival-search:competitive-intel Cursor
/rival-search:fact-check "OpenAI revenue exceeded $2 billion in 2024"
/rival-search:due-diligence Anthropic
/rival-search:trend-intel AI agents for software development
/plugin marketplace add damionrashford/rival-search
/plugin install rival-search@damionrashford-rival-searchgit clone https://github.com/damionrashford/rival-search.git
claude --plugin-dir ./rival-searchThis Claude Code plugin auto-connects to the hosted RivalSearchMCP MCP server. All 10 tools work immediately — no setup, no API keys.
All 10 MCP tools are provided by RivalSearchMCP — zero authentication required.
| Tool | What It Does |
|---|---|
web_search |
Multi-engine search across DuckDuckGo, Yahoo, and Wikipedia |
social_search |
Search Reddit, Hacker News, Dev.to, Product Hunt, Medium |
news_aggregation |
Aggregate news from Google News, DuckDuckGo News, Yahoo News |
github_search |
Search GitHub repositories with rate limiting |
map_website |
Intelligent website exploration and mapping |
content_operations |
Retrieve, analyze, and extract content from any URL |
document_analysis |
Extract text from PDFs, Word docs, and images with OCR |
scientific_research |
Academic papers (arXiv, Semantic Scholar) and datasets (Kaggle, HuggingFace) |
research_topic |
End-to-end research workflow for quick topic analysis |
research_agent |
AI agent with autonomous multi-tool research |
This Claude Code plugin includes 5 agent skills — slash commands that orchestrate MCP tools into multi-step research workflows with structured reports and citations.
| Agent Skill | Description |
|---|---|
/rival-search:research <topic> |
Comprehensive research with academic depth — papers, datasets, implementations |
/rival-search:competitive-intel <company or market> |
Company, product, market, or website intelligence with SWOT analysis |
/rival-search:due-diligence <company or person> |
Full due diligence with risk assessment — works for companies and individuals |
/rival-search:trend-intel <topic> |
News digest + trend trajectory with velocity indicators and maturity assessment |
/rival-search:fact-check <claim> |
Cross-source claim verification with confidence scoring and evidence chains |
This Claude Code plugin includes 6 AI research agents. Each agent has a defined methodology, quality gates, and preloaded agent skills.
| Agent | Specialty |
|---|---|
| research-analyst | Lead researcher — deep multi-source investigation |
| competitive-intel | SWOT analysis, competitor profiling, market positioning |
| fact-checker | Cross-source claim verification, confidence scoring |
| trend-analyst | Trend identification, velocity tracking, trajectory projection |
| content-strategist | Content gap analysis, audience research, brief creation |
| due-diligence | Company/person investigation, risk assessment, red/green flags |
-
.mcp.json— Connects to RivalSearchMCP via MCP athttps://RivalSearchMCP.fastmcp.app/mcp. All 10 MCP tools available automatically. -
skills/— Five agent skill files with step-by-step research workflows, exact MCP tool parameters, and structured output formats. -
agents/— Six AI research agents with full Claude Code frontmatter (tools, agent skills, hooks, memory). Invoked automatically or manually. -
hooks/hooks.json— Portable event hooks for audit logging and output quality validation.
rival-search-plugin/
├── .claude-plugin/
│ ├── plugin.json # Claude Code plugin manifest
│ └── marketplace.json # Marketplace config
├── .mcp.json # MCP server → RivalSearchMCP.fastmcp.app/mcp
├── agents/
│ ├── research-analyst.md
│ ├── competitive-intel.md
│ ├── fact-checker.md
│ ├── trend-analyst.md
│ ├── content-strategist.md
│ └── due-diligence.md
├── skills/
│ ├── research/
│ ├── competitive-intel/
│ ├── due-diligence/
│ ├── trend-intel/
│ └── fact-check/
├── hooks/
│ └── hooks.json
├── CLAUDE.md
├── LICENSE
└── README.md
Does this Claude Code plugin require payment?
No. MIT licensed, powered by RivalSearchMCP. No API keys, no subscriptions, no hidden costs.
Do I need API keys?
No. Install and go. This Claude Code plugin auto-connects to the hosted RivalSearchMCP MCP server.
What's the difference between agent skills and agents?
Agent skills are slash commands you invoke explicitly (/rival-search:research <topic>). Agents are invoked automatically by Claude when your task matches their expertise.
Can I use MCP tools directly without agent skills?
Yes. Agent skills are guided workflows, but you can ask Claude to use any RivalSearchMCP MCP tool directly.
Contributions welcome — new agent skills, AI research agent improvements, or bug fixes.
- Fork the project
- Create your feature branch (
git checkout -b feature/AmazingFeature) - Commit your changes (
git commit -m 'Add AmazingFeature') - Push to the branch (
git push origin feature/AmazingFeature) - Open a Pull Request
- RivalSearchMCP — The MCP server powering this Claude Code plugin
- RivalSearchMCP Docs — Full documentation
- Claude Code — The CLI this plugin extends
MIT — see LICENSE for details.