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/research-deep)<a href="https://agentmods.dev/commands/galbaz1/video-research-mcp/research-deep"><img src="https://agentmods.dev/badge/commands/galbaz1/video-research-mcp/research-deep/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/research-deep"><img src="https://agentmods.dev/badge/commands/galbaz1/video-research-mcp/research-deep.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 | $0.00009 | $0.01257 |
| Opus 5.5 | $0.00004 | $0.00503 |
| Sonnet 5.5 | $0.00002 | $0.00251 |
| Haiku 4.5 | $0.00001 | $0.00126 |
Grade A, and why
research-deep 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 27d 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 — 138 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Deep Research: $ARGUMENTS
Launch the Gemini Deep Research Agent for autonomous web-grounded research (long-running; provider billing applies).
For a shorter model-driven analysis, use
/gr:research. Consult current provider pricing before a paid run.
Phase 0: Load Context
Before interviewing, gather intelligence:
- Prior research: Call
knowledge_search(query="$ARGUMENTS", limit=5)to find existing findings in Weaviate - Check memory: Read files under
<memory-dir>/gr/research/for previous analyses on related topics
Present to the user:
- "I found X prior analyses related to this topic: [summaries]"
- "Here's what we already know: [key findings]"
- "Let me interview you to build a precise research brief."
If no prior context found, proceed directly to the interview.
Phase 1: Research Brief Interview
This is a CHALLENGE-DRIVEN interview. The quality of the brief determines the quality of the research.
Brief Protocol
Use supplied goals, sources, and existing authorization first. If the user requested autonomous work, compile a reversible brief with stated assumptions and proceed within the authorized spend/scope. Ask one consolidated question only for a materially missing decision or required budget authority; do not require an interview or approval again when the session already supplies it.
For interactive brief building, select relevant questions below rather than forcing every round.
Round 1 -- Question Sharpening Restate topic as a precise question. Challenge HARD:
- "This is too broad -- which specific aspect matters for your decision?"
- "What would the ACTIONABLE output look like? A decision memo? Competitive landscape?"
- "What's the actual decision this research needs to inform?"
Round 2 -- Scope Boundaries
- Time period (recent vs historical vs both)
- Domains (academic, industry, regulatory, all)
- Geographic scope if relevant
- What to EXCLUDE (common knowledge, things user already knows)
- Authorized budget and maximum launches/follow-ups
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.
- 27d ago Changed · -4 lines b990aa002e7e
- 1mo ago First seen · 142 lines · 9 tokens per session scan A 5dfc17404af9
research-deep is a command published in the GitHub repository Galbaz1/video-research-mcp (23 stars, last pushed 4d ago), licensed MIT. It adds 9 tokens to every session and 1,257 once invoked, about $0.0000 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
setup-video-vision
Interactive setup wizard for claude-video-vision — configure backend, whisper, frames, and verify dependencies.
eg-new-feature
Build a new feature using the elephant/goldfish workflow — design doc, goldfish design check, implement, review, validate.
eg-fix-bug
Fix a bug using the elephant/goldfish workflow — problem doc, goldfish diagnosis check, failing test, fix, review, validate.
dock-chat
Dock the full conversation to Telegram — drive Claude from your phone.
mem-last
Prints latest memory cards for the current project.
mem-prune
Removes old low-importance memories.