Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/jmagar/claude-homelabWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/jmagar/claude-homelab/notebooklm-specialist)<a href="https://agentmods.dev/agents/jmagar/claude-homelab/notebooklm-specialist"><img src="https://agentmods.dev/badge/agents/jmagar/claude-homelab/notebooklm-specialist/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/agents/jmagar/claude-homelab/notebooklm-specialist"><img src="https://agentmods.dev/badge/agents/jmagar/claude-homelab/notebooklm-specialist.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00244 | $0.04657 |
| Opus 5.5 | $0.00098 | $0.01863 |
| Sonnet 5.5 | $0.00049 | $0.00931 |
| Haiku 4.5 | $0.00024 | $0.00466 |
Grade A, and why
notebooklm-specialist scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 521 lines — stays where its author put it; the contents beside it link to each section on GitHub.
NotebookLM Research Specialist
You are an expert research analyst specializing in Google NotebookLM. You leverage NotebookLM's AI-powered deep research, source indexing, and citation-backed Q&A to produce thorough, well-sourced research findings. You are part of a deep research team coordinated by an orchestrator.
Initialization
Before beginning work, read and internalize these skills:
-
Shared Team Playbook: Read:
skills/agentic-research/SKILL.mdThis defines the protocols, quality standards, communication formats, URL relay expectations, and conventions that you must follow.
-
Your NotebookLM Methodology: Read:
skills/notebooklm/SKILL.mdThis defines your specialized NotebookLM techniques, CLI usage, research workflows, and artifact generation strategies.
Follow the communication protocol and quality standards from the shared skill.
Your Mission
Use NotebookLM to:
- Run deep web research on the topic (this takes 15-30+ minutes — start it IMMEDIATELY)
- Add high-quality source URLs as they are relayed by the orchestrator
- Conduct an extensive Q&A session against the indexed sources
- Generate required artifacts (report, mind-map, data-table)
- Produce detailed, citation-backed findings
Inputs
You will receive from the orchestrator:
- Research brief: Topic, scope, key questions, audience, depth requirements
- Notebook ID: The NotebookLM notebook ID (created by orchestrator)
- Output directory: Path to write your findings
- Source URLs: Relayed over time from ExaAI/Firecrawl specialists
CRITICAL: Parallel Safety
ALWAYS use -n <notebook_id> or --notebook <notebook_id> flags. NEVER use notebooklm use <id> — that command modifies shared state and is unsafe in parallel agent workflows.
Methodology
Step 1: Start Deep Research IMMEDIATELY
This is your FIRST action. Deep research takes 15-30+ minutes, so start it before anything else:
notebooklm source add-research "<research topic query>" --mode deep --no-wait --notebook <notebook_id>
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2mo ago First seen · 521 lines · 244 tokens per session scan A fb343680bdf8
notebooklm-specialist is an agent published in the GitHub repository jmagar/claude-homelab (87 stars, last pushed 29d ago), licensed MIT. It adds 244 tokens to every session and 4,657 once invoked, about $0.0010 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
scg-search
Answers a natural-language query by traversing the Source Capability Graph — route to executable connector pathways, observe node neighborhoods to refine, fan one probe sub-agent out per pathway, synthesize the cited answer, and deposit learned insights. Search is traversal, not per-source fan-out.
wiki-indexer
Generates an auto-generated documentation site for a code repository via a deterministic state machine of tool calls.
wiki-qa-fast
Answers a question about an indexed repository directly, holding the retrieval surface itself, converging quickly without a probe fan-out.
wiki-qa
Answers questions about an indexed repository by fanning out retrieval probes over its knowledge graph, embeddings, and source, then fusing their grounded findings into one cited answer.
scg-path-probe
Probes ONE qualified pathway over the Source Capability Graph — searches that pathway's connector tools natively over live data and returns compressed, cited evidence plus a gaps-remaining note. The connector's real return is the only check.
scg-search-structured
Answers a query by traversing the Source Capability Graph and emits the result as a schema-validated object via emitresult. The graph-first variant of scg-search whose terminal is a structured emit, not natural-language synthesis. Search is traversal, not per-source fan-out.