Deep Research

deep-research

Autonomous multi-step research using a Gemini Deep Research agent — produces thorough, cited reports.

researchgeminiagentcitationsreports

Install

npx skills add sanjay3290/deep-research

Author: sanjay3290
License: Apache-2.0
Updated: February 2026
Type: Single skill
Document: 2.7 KB
Platforms: Claude Code, Claude Ai

View source

SKILL.md

Gemini Deep Research Skill

Run autonomous research tasks that plan, search, read, and synthesize information into comprehensive reports.

Requirements

Python 3.8+
httpx: pip install -r requirements.txt
GEMINI_API_KEY environment variable

Setup

Get a Gemini API key from Google AI Studio
Set the environment variable:
```
export GEMINI_API_KEY=your-api-key-here
```
Or create a .env file in the skill directory.

Usage

Start a research task

python3 scripts/research.py --query "Research the history of Kubernetes"

With structured output format

python3 scripts/research.py --query "Compare Python web frameworks" \
  --format "1. Executive Summary\n2. Comparison Table\n3. Recommendations"

Stream progress in real-time

python3 scripts/research.py --query "Analyze EV battery market" --stream

Start without waiting

python3 scripts/research.py --query "Research topic" --no-wait

Check status of running research

python3 scripts/research.py --status <interaction_id>

Wait for completion

python3 scripts/research.py --wait <interaction_id>

Continue from previous research

python3 scripts/research.py --query "Elaborate on point 2" --continue <interaction_id>

List recent research

python3 scripts/research.py --list

Output Formats

Default: Human-readable markdown report
JSON (--json): Structured data for programmatic use
Raw (--raw): Unprocessed API response

Cost & Time

Metric	Value
Time	2-10 minutes per task
Cost	$2-5 per task (varies by complexity)
Token usage	~250k-900k input, ~60k-80k output

Best Use Cases

Market analysis and competitive landscaping
Technical literature reviews
Due diligence research
Historical research and timelines
Comparative analysis (frameworks, products, technologies)

Workflow

User requests research → Run --query "..."
Inform user of estimated time (2-10 minutes)
Monitor with --stream or poll with --status
Return formatted results
Use --continue for follow-up questions

Exit Codes

0: Success
1: Error (API error, config issue, timeout)
130: Cancelled by user (Ctrl+C)

When to use it

Autonomous, multi-step research that returns a thorough, cited report
Topics deep enough to justify a 2–10 minute agentic run
Producing a sourced briefing you'll hand to a writer or strategist

When not to

Quick factual lookups — the agentic run is overkill (and metered)
Workflows where you can't provision a paid Gemini API key
Tasks needing live site scraping — pair with `firecrawl-cli` for that

Questions

What does it cost to run?: It calls Google's Gemini Deep Research API, which is paid — roughly $2–5 per task, taking 2–10 minutes. You supply the key via GEMINI_API_KEY.
Is it safe to run locally?: Yes — it executes a local Python script that calls only Google's official generativelanguage API. There's no remote code fetch and no shell-out beyond the documented API call.

AI/ML Development

.NET AI

dotnet-ai

AI and ML skills for .NET 8+: technology selection (ML.NET, Microsoft.Extensions.AI, Microsoft Agent Framework, Copilot SDK, ONNX, OllamaSharp), LLM integration, agentic workflows, RAG pipelines, MCP server creation.

claude-codegithub-copilotcursor+1MIT

AI/ML Development

AI Research Skills

ai-research-skills

77 AI research skills across 20 categories covering training, inference, and MLOps workflows.

claude-codeMIT

AI/ML Development

Azure Cognitive Search

azure-cognitive-search

Build search and RAG indexes on Azure Cognitive Search: vector + hybrid search, semantic ranking, indexers, and skillsets. Requires the mcp_microsoftdocs MCP for live docs.

claude-codecodexcursor+3CC-BY-4.0

Agent tooling

Want agents that know your stack?

This index is ours, and so is the tooling behind it. If your team wants agents that understand how your own systems actually work, that is a thing we build.

Contact the dock

One address, read by the people who would do the work. No intake form, no calendar link — just an email.

tphamdn@gmail.com

Other reasons to writeWe reply within two business days