general MCP Server
De-Anthropocentric Research Engine — AI-powered academic research automation with deep literature survey, gap analysis, idea generation, experiment design & execution. Combines iterative deep research, adversarial debate, evolutionary generation, and distributed GPU execution.
Discovered via github-topic:mcp and last synced 4mo ago.
1. Install the package
npm install dare
substitute, combine, adapt, modify, put-other-use, eliminate, reverse
Paper search, Q&A, code exploration (arXiv)
API base URL
SSH username (for remote server targets)
Google Scholar search + web page scraping
[MinerU](https://mineru.net/) API token for PDF → markdown conversion
[Apify](https://console.apify.com/account#/integrations) API token for rag-web-browser
Paper-writing strategy (interface), method-evolve tools (mutate/crossover/evaluate)
Tools
Email for Unpaywall API (polite pool)
Semantic Scholar API — paper lookup, citations, references, recommendations, author info
31 ideation tools (SCAMPER, surgery, cross-domain, perspective, structural)
debate-critic, debate-defender, debate-judge
GPU pod lifecycle management
OpenAI-compatible API key for AI paper reading
SSH hostname/IP (for remote server targets)
LLM micro-agent tools (ideation, debate, insight, method-evolve)
API base URL (optional, for proxy/gateway)
[Semantic Scholar API key](https://www.semanticscholar.org/product/api) (optional — public API works without key at lower rate limits)
Core pipeline (intake, lit-survey, round, scholar SOPs, web SOPs)
Academic paper pipeline — search, enrich, fetch, read, reference
Model ID (default: `claude-sonnet-4-20250514`)
INSIGHT 7-step pipeline, multiagent debate, quality-diversity filtering
root-cause-drilling, stakeholder-mapping, tension-mining, question-reformulation, abstraction-laddering, assumption-audit, validation
Web search API
Cache directory, shared with dare-scholar (**must be an absolute path**)
RunPod API key (for GPU pod targets)
Tools
API key for LLM completions (Anthropic or compatible proxy)
Component
### 📁 Monorepo Structure ```text dare/ ├── packages/ │ ├── agents/ # dare-agents MCP — 49 LLM micro-agent tools (72 tests) │ ├── scholar/ # dare-scholar MCP — academic paper pipeline (5 tools) │ ├── ss/ # dare-ss MCP — Semantic Scholar API (8 tools, 75 tests) │ ├── web/ # dare-web MCP — web page fetching & caching (2 tools) │ └── session/ # dare-session — pod provisioning scripts ├── skills/ │ ├── dare/ # /dare meta-strategy (entry point) │ ├── strategy/ # 8 strategies (lit-survey, gap-analysis, insight, ...) │ ├── tactic/ # 15 tactics (debate, scamper, surgery, ...) │ └── sop/ # 60 SOPs (one per dare-agents tool) ├── package.json # Root workspace config └── .mcp.json # MCP server configuration (gitignored) ``` ### 🔌 MCP Servers
Description
Model name for paper reading agent
Hugging Face token (passed to pod for model downloads)