Atlas / MCP servers / yoloshii / Gigaxity Deep Research

Gigaxity Deep ResearchSAFE

Name: Gigaxity Deep Research
Author: yoloshii

mcp/yoloshii/gigaxity-deep-research

MCP server by yoloshii · agentic ai · agentic search · ai agents · Audited 2026-10-08

Open-source deep research MCP. Qwen3-30B-A3B-Thinking via OpenRouter, cited web synthesis for Claude Code, Codex, Cursor, Hermes and any MCP-compatible agent.

Verdict

SAFE

Grade

Trust score

88 /100

Exposed tools

25 24r · 1w · 0d

Transport

stdio

License

MIT

Stars

Overview

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

Gigaxity Deep Research — Open-source deep research MCP server for Claude Code, Codex, Cursor, Hermes, and any MCP-capable agent

Open-source deep research MCP server for Claude Code, Hermes, Cursor, and any MCP-compatible agent. Qwen3-30B-A3B-Thinking via OpenRouter plus multi-source web synthesis with citations.

Gigaxity Deep Research is a multi-source synthesis pipeline — seven MCP tools (three primitives search/vertical_search/research plus four deep-research tools ask/discover/synthesize/reason) with a matching FastAPI REST surface, fronting parallel multi-source search, RRF fusion, citation binding, and contradiction detection. The synthesis stage runs against any OpenAI-compatible chat-completions model; the recommended default is Alibaba's Qwen3-30B-A3B-Thinking, a reasoning-tuned 30B-A3B MoE model, but DeepSeek-R1, Qwen-QwQ, Llama 3.x, and hosted-aggregator endpoints (OpenRouter and the like) all work — pick any model your endpoint serves. The search layer pulls from a "Triple Stack" of complementary MCPs (Context7, Exa, Jina) alongside SearXNG, Tavily, LinkUp, Brave, and Parallel connectors. A bundled `gptr-mcp` companion — the MCP shim around GPT Researcher — adds Reddit, X, and YouTube as social-first sources.

If you want to run the synthesis model on your own hardware, the local-inference branch swaps OpenRouter for any OpenAI-compatible endpoint (vLLM, SGLang, or llama.cpp). The search-MCP layer is priced separately by each provider. See [docs/guides/free-tier-strategy.md](docs/guides/free-tie

Read from source at commit 2d2a884c2fc0OBSERVED · 2026-10-08

Connect

Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.

claude-code

claude mcp add gigaxity-deep-research --env BRIGHTDATA_API_TOKEN=${BRIGHTDATA_API_TOKEN} --env EXA_API_KEY=${EXA_API_KEY} --env JINA_API_KEY=${JINA_API_KEY} --env JINA_GHOST_KEY=${JINA_GHOST_KEY} -- uvx gigaxity-deep-research

claude-desktop

{
  "mcpServers": {
    "gigaxity-deep-research": {
      "command": "uvx",
      "args": [
        "gigaxity-deep-research"
      ],
      "env": {
        "BRIGHTDATA_API_TOKEN": "${BRIGHTDATA_API_TOKEN}",
        "EXA_API_KEY": "${EXA_API_KEY}",
        "JINA_API_KEY": "${JINA_API_KEY}",
        "JINA_GHOST_KEY": "${JINA_GHOST_KEY}"
      }
    }
  }
}

Exposed tools (25)

24 read · 1 write · 0 destructive.

Tool	Risk	Description
`capture_screenshot_url`	read	Capture a screenshot of a webpage and return the hosted image URL.
`classify_text`	read	Zero-shot classify texts into supplied labels.
`deduplicate_strings`	read	Pick a semantically diverse subset, dropping near-duplicates.
`exa_answer`	read	Fast factual answer with citations. 1-2s, 94% SimpleQA accuracy.
`exa_answer_detailed`	read	Detailed factual answer with full source text. For when you need more context.
`extract_pdf`	read	Extract layout elements (figures, tables, equations) from a PDF.
`guess_datetime_url`	read	Estimate when a page was published, for freshness and credibility checks.
`parallel_read_url`	read	Read several URLs concurrently. Preferred over repeated read_url calls.
`parallel_search_arxiv`	read	Search arXiv for several queries concurrently.
`parallel_search_ssrn`	read	Search SSRN for several queries concurrently.
`parallel_search_web`	write	Run several web searches concurrently. Use for query variants in one pass.
`primer`	read	Report current UTC time and this server
`read_url`	read	Read a webpage or PDF and return clean markdown.
`scrape_as_markdown`	read	Scrape a blocked URL as markdown using Brightdata Web Unlocker.
`search`	read	Multi-source search with RRF (Reciprocal Rank Fusion).
`search_arxiv`	read	Search arXiv for academic papers.
`search_bibtex`	read	Find BibTeX citation entries for a paper.
`search_images`	read	Search for images on the web.
`search_jina_blog`	read	Search Jina AI
`search_ssrn`	read	Search SSRN for economics, law, and finance working papers.
`search_web`	read	Search the live web. Returns titles, URLs and descriptions.
`show_api_key`	read	Show the masked API key this server is using, and its wallet balance.
`sort_by_relevance`	read	Rerank documents by semantic relevance to a query.
`synthesize`	read	It makes no LLM call. A reporter here would emit an opening tick and then
`vertical_search`	read	Search ONE SearXNG category. No LLM call and no search-API quota.

Trust audit

SAFEgrade B · trust 88/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

Layer	What it checks	Result
L0	Provenance & inventory	PASS
L1	Static analysis of the code	PASS
L2	Instruction surface (what it tells the agent)	PASS
L3	Class-specific surface	PASS
L4	Behavioural (sandbox)	SKIPPED

What the source does

Filesystem: none-observed
Network: declared (7 observation(s))
Shell: none-observed
Dependencies: not all pinned
Secrets in source: none-found

Findings (15)

LOWInsecure crypto · crypto.weak_hash · CWE-327, CWE-338

src/connectors/brave.py:81

source_id = f"br_{hashlib.md5(url.encode()).hexdigest()[:8]}"

LOWInsecure crypto · crypto.weak_hash · CWE-327, CWE-338

src/connectors/linkup.py:52

source_id = f"lu_{hashlib.md5(url.encode()).hexdigest()[:8]}"

LOWInsecure crypto · crypto.weak_hash · CWE-327, CWE-338

src/connectors/parallel.py:113

source_id = f"px_{hashlib.md5(url.encode()).hexdigest()[:8]}"

LOWInsecure crypto · crypto.weak_hash · CWE-327, CWE-338

src/connectors/searxng.py:314

source_id = f"sx_{hashlib.md5(url.encode()).hexdigest()[:8]}"

LOWInsecure crypto · crypto.weak_hash · CWE-327, CWE-338

src/connectors/tavily.py:53

source_id = f"tv_{hashlib.md5(url.encode()).hexdigest()[:8]}"