video-research-mcp

video-research-mcp is a Claude Code plugin that installs and configures MCP servers to enable video understanding/analysis, deep research with evidence grading, academic/paper search and extraction, web search with citations, and an optional knowledge store (Weaviate) for persistent semantic recall. It also describes standalone MCP server usage and related explainer/video/scene-generation MCP servers.

Evaluated Mar 30, 2026 (21d ago)

Repo ↗ Ai Ml mcp claude-code video-analysis deep-research weaviate research-citations knowledge-graph media-production

⚙ Agent Friendliness

/ 100

Can an agent use this?

🔒 Security

/ 100

Is it safe for agents?

⚡ Reliability

/ 100

Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality

Documentation

Error Messages

Auth Simplicity

Rate Limits

🔒 Security

TLS Enforcement

Auth Strength

Scope Granularity

Dep. Hygiene

Secret Handling

Security is largely based on external provider API keys via environment variables. The README specifies environment-variable configuration but does not describe secure storage, redaction, logging behavior, or transport guarantees beyond typical HTTPS usage. No evidence of fine-grained access controls/scope granularity is provided. Multiple integrations (Gemini, Weaviate, optional providers) increase the risk surface; users should ensure keys are not logged and understand what data is sent to third parties (e.g., video uploads >20MB to Gemini File API).

⚡ Reliability

Uptime/SLA

Version Stability

Breaking Changes

Error Recovery

Best When

You want an MCP-capable agent workflow that combines video analysis + citation-grounded research and optionally persistent semantic memory using Weaviate.

Avoid When

You need formal API contracts (OpenAPI) or strong, documented reliability/error-handling semantics at the protocol level; also avoid when secrets/data cannot be sent to external AI providers or stored in third-party services.

Use Cases

• Analyze local video files (e.g., meetings), extracting timestamps/decisions/action items and producing concept maps
• Analyze YouTube tutorials with timestamps and sentiment/comment analysis
• Run deep research workflows with evidence-tier labeling and source-grounded claims
• Ingest and query past analyses via a Weaviate-backed knowledge store (semantic recall)
• Analyze documents (URLs/files/directories) to extract entities/relationships and synthesize across sources
• Search the web and retrieve cited results
• Generate explainer video projects and media production pipelines via additional MCP servers (video-explainer-mcp / video-agent-mcp)

Not For

• Highly regulated environments requiring strict data residency/compliance guarantees (not evidenced in provided materials)
• Environments where you cannot provide and secure multiple third-party API keys (Gemini, optional ElevenLabs/OpenAI/Weaviate/Cohere/Semantic Scholar)
• Use cases requiring an officially documented public REST/GraphQL API for direct integration beyond MCP/Claude-Code workflows
• Guarantees of deterministic outputs or strict SLAs for research/media generation (not documented)

Interface

REST API

GraphQL

gRPC

MCP Server

Yes

SDK

Webhooks

Authentication

Methods: API keys via environment variables (GEMINI_API_KEY; optional WEAVIATE_API_KEY, COHERE_API_KEY, ELEVENLABS_API_KEY, OPENAI_API_KEY, YOUTUBE_API_KEY, S2_API_KEY) MCP client authentication is effectively inherited from the MCP server process environment

OAuth: No Scopes: No

Authentication is configured via environment variables for upstream services; no OAuth flow or explicit fine-grained scopes are described in the provided README.

Pricing

Model: Gemini/other upstream providers (e.g., Weaviate/Op

Free tier: No

Requires CC: No

No pricing or free tier information for the package itself is provided; costs depend on enabled third-party APIs (Gemini, ElevenLabs/OpenAI, Weaviate, etc.).

Agent Metadata

Pagination

none

Idempotent

False

Retry Guidance

Not documented

Known Gotchas

⚠ Large local files (>20MB) are uploaded to Gemini File API for context-caching; agents may incur upload latency/cost and should avoid repeated uploads where possible.
⚠ Weaviate-backed semantic recall is optional; without configuration, recall falls back to keyword grep, which may reduce answer quality.
⚠ Some advanced capabilities depend on optional keys/services (YouTube Data API, Semantic Scholar, Cohere reranker, ElevenLabs/OpenAI/Sora, MLflow tracing).
⚠ The package relies on Claude Code plugin installation/assets plus MCP servers; integration may vary by client (Claude Code vs generic MCP client).

Alternatives

Other MCP servers or SDKs for video understanding and web research (separate tools for each capability) Weaviate + separate RAG pipelines (custom document processing and retrieval) Commercial all-in-one research/video automation platforms (if available) Claude Code plugins focused solely on research or solely on media generation

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for video-research-mcp.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

API endpoint ↗ Agent guide ↗ Report inaccuracy

Scores are editorial opinions as of 2026-03-30.