Deep Research

1---
2name: deep-research
3description: "Execute autonomous multi-step research using Google Gemini Deep Research Agent. Use for: market analysis, competitive landscaping, literature reviews, technical research, due diligence. Takes 2-10 minutes but produces detailed, cited reports. Costs $2-5 per task."
4license: Apache-2.0
5metadata:
6  author: sanjay3290
7  version: "1.0"
8---
9 
10# Gemini Deep Research Skill
11 
12Run autonomous research tasks that plan, search, read, and synthesize information into comprehensive reports.
13 
14## Requirements
15 
16- Python 3.8+
17- httpx: `pip install -r requirements.txt`
18- GEMINI_API_KEY environment variable
19 
20## Setup
21 
221. Get a Gemini API key from [Google AI Studio](https://aistudio.google.com/)
232. Set the environment variable:
24   ```bash
25   export GEMINI_API_KEY=your-api-key-here
26   ```
27   Or create a `.env` file in the skill directory.
28 
29## Usage
30 
31### Start a research task
32```bash
33python3 scripts/research.py --query "Research the history of Kubernetes"
34```
35 
36### With structured output format
37```bash
38python3 scripts/research.py --query "Compare Python web frameworks" \
39  --format "1. Executive Summary\n2. Comparison Table\n3. Recommendations"
40```
41 
42### Stream progress in real-time
43```bash
44python3 scripts/research.py --query "Analyze EV battery market" --stream
45```
46 
47### Start without waiting
48```bash
49python3 scripts/research.py --query "Research topic" --no-wait
50```
51 
52### Check status of running research
53```bash
54python3 scripts/research.py --status <interaction_id>
55```
56 
57### Wait for completion
58```bash
59python3 scripts/research.py --wait <interaction_id>
60```
61 
62### Continue from previous research
63```bash
64python3 scripts/research.py --query "Elaborate on point 2" --continue <interaction_id>
65```
66 
67### List recent research
68```bash
69python3 scripts/research.py --list
70```
71 
72## Output Formats
73 
74- **Default**: Human-readable markdown report
75- **JSON** (`--json`): Structured data for programmatic use
76- **Raw** (`--raw`): Unprocessed API response
77 
78## Cost & Time
79 
80| Metric | Value |
81|--------|-------|
82| Time | 2-10 minutes per task |
83| Cost | $2-5 per task (varies by complexity) |
84| Token usage | ~250k-900k input, ~60k-80k output |
85 
86## Best Use Cases
87 
88- Market analysis and competitive landscaping
89- Technical literature reviews
90- Due diligence research
91- Historical research and timelines
92- Comparative analysis (frameworks, products, technologies)
93 
94## Workflow
95 
961. User requests research → Run `--query "..."`
972. Inform user of estimated time (2-10 minutes)
983. Monitor with `--stream` or poll with `--status`
994. Return formatted results
1005. Use `--continue` for follow-up questions
101 
102## Exit Codes
103 
104- **0**: Success
105- **1**: Error (API error, config issue, timeout)
106- **130**: Cancelled by user (Ctrl+C)
107