librarium

分类: 调研与分析 | 上传者: jkudish | 下载: 0 | 版本: v1.0（最新）

使用 librarium CLI 运行多提供商深度研究查询

更新日志: Source: GitHub https://github.com/jkudish/librarium

目录结构

当前层级: tree/main/

📁 .github/
- 📁 ISSUE_TEMPLATE/
  - 📄 bug_report.yml 2.3 KB
  - 📄 config.yml 357 B
  - 📄 feature_request.yml 1.3 KB
  - 📄 new_provider.yml 1.6 KB
- 📁 scripts/
  - 📄 parse-changelog.sh 4.9 KB
- 📁 workflows/
  - 📄 ci.yml 2.3 KB
  - 📄 release.yml 12.5 KB
- 📄 FUNDING.yml 18 B
- 📄 pull_request_template.md 611 B
📁 art/
- 📄 gh-og.png 8.6 MB
📁 docs/
- 📄 provider-development.md 5.1 KB
📁 scripts/
- 📁 custom-providers/
  - 📄 wikipedia-provider.mjs 3.2 KB
- 📄 build-sea.mjs 2.9 KB
- 📄 install.sh 3.3 KB
📁 src/
- 📁 adapters/
  - 📄 base.ts 2.6 KB
  - 📄 brave-answers.ts 3.8 KB
  - 📄 brave-search.ts 3.3 KB
  - 📄 custom.ts 15.3 KB
  - 📄 exa.ts 3.8 KB
  - 📄 firecrawl-search.ts 3.7 KB
  - 📄 gemini-deep.ts 5.6 KB
  - 📄 index.ts 4.6 KB
  - 📄 jina-search.ts 3.6 KB
  - 📄 kagi-fastgpt.ts 4.0 KB
  - 📄 openai-deep-base.ts 7.2 KB
  - 📄 openai-deep-o3.ts 324 B
  - 📄 openai-deep.ts 330 B
  - 📄 perplexity-advanced-deep.ts 535 B
  - 📄 perplexity-agent-base.ts 5.8 KB
  - 📄 perplexity-deep-research.ts 494 B
  - 📄 perplexity-search.ts 3.5 KB
  - 📄 perplexity-sonar-deep.ts 4.5 KB
  - 📄 perplexity-sonar-pro.ts 3.3 KB
  - 📄 searchapi.ts 3.6 KB
  - 📄 serpapi.ts 3.5 KB
  - 📄 tavily.ts 3.8 KB
  - 📄 you-research.ts 3.3 KB
- 📁 commands/
  - 📄 cleanup.ts 4.7 KB
  - 📄 config.ts 3.0 KB
  - 📄 doctor.ts 4.6 KB
  - 📄 groups.ts 2.3 KB
  - 📄 init.ts 3.7 KB
  - 📄 install-skill.ts 2.7 KB
  - 📄 ls.ts 2.9 KB
  - 📄 run.ts 8.4 KB
  - 📄 status.ts 10.0 KB
  - 📄 upgrade.ts 4.6 KB
- 📁 core/
  - 📄 async-manager.ts 1.9 KB
  - 📄 config.ts 7.9 KB
  - 📄 dispatcher.ts 11.5 KB
  - 📄 fs-utils.ts 566 B
  - 📄 http-client.ts 3.7 KB
  - 📄 install-method.ts 1.8 KB
  - 📄 normalizer.ts 2.3 KB
  - 📄 prompt-builder.ts 1.1 KB
  - 📄 synthesis.ts 3.1 KB
- 📄 cli.ts 1.3 KB
- 📄 constants.ts 4.7 KB
- 📄 types.ts 6.1 KB
📁 tests/
- 📁 adapters/
  - 📄 base-error-helpers.test.ts 3.7 KB
- 📁 commands/
  - 📄 install-skill.test.ts 1.2 KB
  - 📄 upgrade.test.ts 1.6 KB
- 📁 integration/
  - 📄 cli.test.ts 2.6 KB
- 📄 config.test.ts 11.5 KB
- 📄 custom-providers.test.ts 10.0 KB
- 📄 dispatcher-fallback.test.ts 11.5 KB
- 📄 http-client.test.ts 7.3 KB
- 📄 install-method.test.ts 2.3 KB
- 📄 normalizer.test.ts 3.4 KB
- 📄 registry.test.ts 5.7 KB
- 📄 slug.test.ts 1.2 KB
📄 .gitattributes 19 B
📄 .gitignore 49 B
📄 biome.json 596 B
📄 CHANGELOG.md 4.8 KB
📄 LICENSE 1.0 KB
📄 package-lock.json 124.8 KB
📄 package.json 1.5 KB
📄 README.md 18.9 KB
📄 SKILL.md 3.3 KB
📄 tsconfig.json 454 B
📄 tsup.config.ts 425 B
📄 vitest.config.ts 216 B
📄 vitest.integration.config.ts 189 B

SKILL.md

---
description: Run multi-provider deep research queries using the librarium CLI
triggers:
  - /librarium
  - /research
  - /deep-research
  - deep research
  - multi-provider search
---

# Librarium -- Multi-Provider Deep Research

Run research queries across 10 search and deep-research APIs in parallel, collect results, deduplicate sources, and produce structured output.

## Prerequisites

- `librarium` CLI installed (`npm install -g librarium`)
- API keys configured (`librarium init --auto`)
- Binary at: `librarium` (or `npx librarium`)

## 7-Phase Research Workflow

### Phase 1: Query Analysis
Analyze the user's research question. Determine:
- Is this a technical, business, or general knowledge query?
- Which provider group is best suited? (`quick` for fast answers, `deep` for thorough research, `comprehensive` for important decisions, `all` for maximum coverage)
- What execution mode? (`sync` for quick queries, `mixed` for deep research)

### Phase 2: Provider Selection
Select providers based on query type:
- **Technical queries**: Use `comprehensive` group (deep research + AI-grounded)
- **Quick facts**: Use `quick` group (AI-grounded only, fast)
- **Competitive research**: Use `all` group (maximum coverage)
- **Specific provider**: Use `--providers` flag

### Phase 3: Dispatch
Run the query:
```bash
librarium run "your query here" --group <group> [--mode mixed]
```

### Phase 4: Monitor Async Tasks
If deep-research providers were used in async mode:
```bash
librarium status --wait
```

### Phase 5: Retrieve Results
Once complete, async results can be retrieved:
```bash
librarium status --retrieve
```

### Phase 6: Analyze Output
Read the output files:
1. `summary.md` -- Overall research summary with statistics
2. `sources.json` -- Deduplicated citations ranked by frequency
3. Individual `{provider}.md` files for detailed per-provider results
4. `run.json` -- Machine-readable manifest

### Phase 7: Synthesize
Combine findings from multiple providers into a coherent answer. Cross-reference sources that appear across multiple providers (higher citation count = higher confidence).

## Key Commands

| Command | Purpose |
|---------|---------|
| `librarium run <query>` | Run research query |
| `librarium run <query> --group quick` | Fast AI-grounded search |
| `librarium run <query> --group deep` | Deep research (async) |
| `librarium run <query> --group all` | All providers |
| `librarium status` | Check async tasks |
| `librarium status --wait --retrieve` | Wait and fetch async results |
| `librarium ls` | List providers and status |
| `librarium doctor` | Health check providers |
| `librarium config` | Show resolved config |

## Provider Tiers

| Tier | Providers | Speed | Depth |
|------|-----------|-------|-------|
| deep-research | perplexity-sonar-deep, perplexity-deep-research, perplexity-advanced-deep, openai-deep, openai-deep-o3, gemini-deep | Minutes | Comprehensive |
| ai-grounded | perplexity-sonar-pro, brave-answers, exa, you-research, kagi-fastgpt | Seconds | Good |
| raw-search | perplexity-search, brave-search, jina-search, searchapi, serpapi, tavily | Fast | Links only |

## Output Structure

```
./agents/librarium/{timestamp}-{slug}/
  prompt.md, run.json, summary.md, sources.json
  {provider}.md, {provider}.meta.json
  async-tasks.json (if applicable)
```

登录后下载/点赞/收藏 ❤ 117 | ★ 0

librarium

目录结构

SKILL.md

举报内容

提示