notebooklm-research
VerifiedAutomate source-grounded research with Google NotebookLM. Create notebooks from URLs, text, or local files; ask cited questions; run fast or deep web research; create articles and social drafts; and generate or download audio, video, cinematic video, slides, reports, study guides, quizzes, flashcards, mind maps, infographics, and data tables. Use when a user asks for NotebookLM, cited source analysis, research-to-content workflows, podcasts, slides, study material, artifact generation, RSS digests, or trend research.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 7.3 KB |
| examples/research-to-article/README.md | markdown | 1.9 KB |
| examples/research-to-threads/README.md | markdown | 1.4 KB |
| examples/trend-to-content/README.md | markdown | 2.0 KB |
| references/api_surface.md | markdown | 6.4 KB |
| references/output_formats.md | markdown | 3.2 KB |
| references/pipeline_recipes.md | markdown | 3.5 KB |
| scripts/__init__.py | script | 0 B |
| scripts/auth_helper.pyexecutable | script | 6.0 KB |
| scripts/common.py | script | 26.5 KB |
| scripts/make_video.shexecutable | script | 3.7 KB |
| scripts/notebooklm_client.pyexecutable | script | 20.2 KB |
| scripts/pipeline.pyexecutable | script | 26.1 KB |
| scripts/skill_installer.pyexecutable | script | 4.5 KB |
references/pipeline_recipes.md
View on GitHub# Pipeline recipes
These examples use only commands implemented by version 1.4.x. All results are JSON;
use `jq` or another JSON parser instead of scraping stderr.
## Sources to article
```bash
notebooklm-pipeline research-to-article \
--sources https://example.com/official https://example.com/independent \
--files ./paper.pdf \
--title "Evidence review" \
--language en \
--audience "technical decision makers" \
--tone "precise and skeptical" > article-result.json
```
The workflow asks five evidence questions and then requests a grounded article. Check
`.status`, `.source_summary`, every `.research_findings[].status`, and
`.article.status` before using the draft.
## Sources to social drafts
```bash
notebooklm-pipeline research-to-social \
--sources https://example.com/a https://example.com/b \
--platform threads \
--language zh-TW \
--variants 3 > social-result.json
```
Supported platforms are `threads`, `twitter`, `linkedin`, `instagram`, and `generic`.
The result is a draft for review; this project has no publishing command.
## RSS digest and Q&A
```bash
notebooklm-pipeline batch-digest \
--rss https://example.com/feed.xml \
--title "Weekly digest" \
--max-entries 20 \
--qa-count 5 \
--rss-timeout 30 \
--max-feed-bytes 5000000 > digest-result.json
```
The feed fetch follows HTTP redirects, enforces time and byte limits, parses RSS/Atom,
and ingests individual article URLs. Empty feeds and zero successful sources fail
before question generation.
## Trend research
Install `trend-pulse`, or point at its absolute executable:
```bash
export TREND_PULSE_CMD=/absolute/path/to/trend-pulse
notebooklm-pipeline trend-to-content \
--geo TW \
--count 5 \
--platform threads \
--language zh-TW \
--research-mode deep > trends-result.json
```
For each trend the pipeline ingests usable URLs (or descriptive text), runs NotebookLM
web research, imports results, and creates a grounded draft. One topic failing does not
erase successful topics; expect `partial` when appropriate.
## Batch artifact generation
```bash
notebooklm-pipeline generate-all \
--files ./paper.pdf \
--types audio slides report mind-map data-table \
--language en \
--output-dir ./output \
--artifact-concurrency 2 \
--force > artifacts-result.json
```
Generation jobs start first and then finish with bounded concurrency. Use
`--no-wait --no-download` to enqueue jobs without waiting. `--download` cannot be
combined with `--no-wait`.
## Manual deep research plus podcast
```bash
notebooklm-skill create \
--title "AI coding evidence" \
--sources https://example.com/study https://example.com/documentation \
--strict
notebooklm-skill research \
--notebook "AI coding evidence" \
--query "Independent productivity and quality findings" \
--mode deep --max-sources 10
notebooklm-skill podcast \
--notebook "AI coding evidence" \
--audio-format debate \
--audio-length long \
--instructions "Contrast supportive and skeptical evidence" \
--output ./output/podcast.m4a
```
For long jobs, enqueue with `generate --type audio --no-wait`, inspect with
`list-artifacts`, then download by exact artifact ID.
## Automation rules
1. Select the profile explicitly in unattended jobs.
2. Parse JSON and stop on nonzero exits.
3. Treat `partial` as review-required, not success.
4. Retain citation metadata alongside generated prose.
5. Limit concurrency and avoid retry storms after quota errors.
6. Use notebook/artifact IDs rather than ambiguous titles.
7. Never add `--force` or deletion confirmation unless overwrite/delete is intended.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| deprecation-and-migrationaddyosmani/agent-skillsManages deprecation and migration. Use when removing old systems, APIs, or features. Use when migrating users from one implementation to another. Use when migrating a database schema in production, such as renaming or dropping a column without downtime (expand/contract). Use when deciding whether to maintain or sunset existing code. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| performance-optimizationaddyosmani/agent-skillsOptimizes application performance across frontend, backend, queries, and databases. Use when performance requirements exist, when you suspect performance regressions, when Core Web Vitals or load times need improvement, when N+1 query patterns need fixing, or when profiling reveals bottlenecks. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| agent-cost-reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| agent-cost-reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |