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/Galbaz1/video-research-mcpWrote 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/commands/galbaz1/video-research-mcp/explain-video)<a href="https://agentmods.dev/commands/galbaz1/video-research-mcp/explain-video"><img src="https://agentmods.dev/badge/commands/galbaz1/video-research-mcp/explain-video/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/commands/galbaz1/video-research-mcp/explain-video"><img src="https://agentmods.dev/badge/commands/galbaz1/video-research-mcp/explain-video.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 | $0.00017 | $0.02412 |
| Opus 5.5 | $0.00007 | $0.00965 |
| Sonnet 5.5 | $0.00003 | $0.00482 |
| Haiku 4.5 | $0.00002 | $0.00241 |
Grade A, and why
explain-video 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 22d 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 — 188 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Explain Video: $ARGUMENTS
Research a topic or analyze content, then synthesize it into an explainer video.
Parse Arguments
$ARGUMENTS should contain:
- A URL (YouTube, webpage) or topic to research
- A project ID for the explainer video
If only one argument, use it as both the research subject and derive the project ID.
Phase 1: Research & Analysis
Based on input type:
YouTube URL: Call video_analyze(url, instruction="Extract key concepts, structure, and talking points for creating an educational explainer video")
Webpage URL or supplied evidence: Prepare research_execute with mode
retrieval, supplied, or hybrid, explicit permitted URLs/domains or the
original EvidencePacket and source_root, and dry_run: true. Execution needs
actual source-access/submission authority and a shared call/token/source allowance.
Use only returned retained sources; keep rejected and failed branches visible.
Topic text: Select the route explicitly. Use research_execute in
model_only mode for an unverified model draft; label its findings accordingly.
For fresh external facts, use authorized research_web and poll the exact job,
or provide permitted seed URLs to the bounded retrieval route. Hosted research
does not expose enforceable internal search/token/USD ceilings. A citation URI
from hosted research is a lead until the original source bytes are retained.
Never describe research_deep as observed retrieval: its three model calls are
model-only synthesis, and model-written tiers do not establish CONFIRMED facts.
Use the MCP tool's typed request schema and returned recovery/source report as
the executable contract. Repository documentation at
docs/integrations/grounded-research.md is an optional checkout-only reference.
A dry plan and the executed request have distinct run identities. Save
the executed run_id; replay its same request to inspect an accepted terminal
packet or an ambiguous failure without automatically resubmitting paid work.
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.
- 22d ago Changed · +127 lines d1186b0c6200
- 24d ago Changed · -1 lines eff065c7ee24
- 1mo ago First seen · 62 lines · 17 tokens per session scan A 515727d6d7c1
explain-video is a command published in the GitHub repository Galbaz1/video-research-mcp (23 stars, last pushed yesterday), licensed MIT. It adds 17 tokens to every session and 2,412 once invoked, about $0.0001 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 commands, from other repositories
zooza-setup
One-time Zooza setup — teaches Claude your business vocabulary so it understands your terms in every future session.
setup-video-vision
Interactive setup wizard for claude-video-vision — configure backend, whisper, frames, and verify dependencies.
grill-me
Run an interactive one-question-at-a-time oral drill for research defence or active-recall study, escalating around weak answers and ending with a study sheet. Use when the user asks to be grilled, quizzed, or prepared for a viva, job talk, seminar, or exam. Not for a written critique; use $devils-advocate or a review…
init-project-course
Bootstrap a university course or module folder with the canonical teaching structure, guidance, and starter files. Use when creating or migrating a course workspace. Not for checking an existing course without modifying it; use $audit-project-course.
specialization
Train agents to become experts in specific domains for better performance.
eg-fix-bug
Fix a bug using the elephant/goldfish workflow — problem doc, goldfish diagnosis check, failing test, fix, review, validate.