# Run Structured Deep Research with Gemini

Complex research can require many searches, source checks, and synthesis steps. This skill documents a Gemini workflow for starting, monitoring, continuing, and formatting research tasks.

## Install

```bash
npx skillstore add sickn33/deep-research
```

## Metadata

- Status: approved
- Slug: sickn33-deep-research
- Skillstore revision: r2
- Version status: missing
- Tree hash: 2aa1cfdef329a8137c6d961fdb7b8d33ea5158f4050eb2a764379fbcd1294843
- Author: sickn33
- GitHub username: sickn33
- License: MIT
- Repository: https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/deep-research
- Ref: f9e2c34b4f19c7f3e6b0a1e93227b5f77cc12526
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands, network, env\_access
- Quality score: 78
- Quality tier: bronze
- Public page: https://skillstore.pages.dev/skills/sickn33-deep-research
- Manifest: https://skillstore.pages.dev/api/skills/sickn33-deep-research/manifest

## Capabilities

- Documents research workflows for market analysis, literature reviews, technical research, competitive analysis, and due diligence.
- Provides command patterns for starting, streaming, polling, waiting for, continuing, and listing Gemini research tasks.
- Shows how to request a custom report structure for comparative analysis.
- Describes human-readable Markdown, structured JSON, and raw response output modes.
- Explains expected task duration, estimated cost, token usage, workflow steps, and exit codes.

## Use Cases

- Assess a market: Map competitors, market growth, regulation, and recent forecasts before planning a market entry.
- Review academic evidence: Build a cited literature overview, compare findings, and identify unresolved questions before deeper source review.
- Compare technologies: Compare technologies using evidence about capabilities, adoption, maintenance, costs, and operational risks.

## Prompt Templates

### Research a topic

```
Research [topic]. Provide a concise overview, key dates, major sources, and unresolved questions. Cite every factual claim.
```

### Compare two options

```
Compare [option A] and [option B] for [context]. Evaluate cost, capabilities, risks, and adoption evidence. Finish with a recommendation.
```

### Analyze a market

```
Analyze [market] in [region] from [start year] to [end year]. Cover size, growth, competitors, regulation, and cited forecasts.
```

### Test a complex decision

```
Investigate [decision] using primary and recent sources. Separate evidence from inference, challenge assumptions, identify conflicts, and present confidence-rated conclusions.
```

## Limitations

- The package contains only SKILL.md; the referenced research script and requirements file are not included.
- Research execution requires a separate Gemini API key, compatible implementation, internet access, and paid service availability.
- Cost, duration, and token estimates are approximate and can vary by task.
- Generated research can contain outdated, incomplete, or incorrect claims and requires source verification.

## Best Practices

- Define the topic, geography, time range, audience, and decision before starting research.
- Request primary and recent sources, then verify important claims against the cited material.
- Use continuation requests to challenge assumptions, resolve conflicting evidence, and deepen weak sections.

## Anti Patterns

- Do not submit broad topics without scope, decision criteria, or a required output structure.
- Do not treat generated citations or conclusions as verified without opening and checking the sources.
- Do not place API keys in prompts, command history, reports, or tracked files.

## Security Audit

- Audited at: 2026-07-23T21:46:36.896\+00:00
- Summary: All 30 static findings are false positives caused by Markdown backticks, code fences, setup links, API-key placeholders, and exit-code prose. The package contains only SKILL.md and no executable implementation, so no command execution, secret access, reconnaissance, or exfiltration behavior is present. No prompt injection or net-new semantic security issue was found.

## Stats

- Views: 104
- Downloads: 21
- Favorites: 0
- Popularity score: 0