MCP YouTube Transcript
An MCP server retrieving transcripts of YouTube videos
Description from the maker (original English)
MCP server for Affinity CRM - entity search, relationship intelligence, and workflow management
These three are never merged into one score: opinions, measurements and popularity are different things.
We have not written a plain-language summary for this one yet. Below are the facts we read from the registry.
Pick the AI app you use, then copy the settings.
AFFINITY_API_KEYValues marked secret are keys to your account. Never share them or paste them into a chat.
Which app do you use?
Pick your app above. We remember it, so next time the button says “Install in …”.
Download the maker’s .mcpb file
The link goes to the maker’s own file; we do not host or repackage it. Open the file and Claude Desktop installs it.
Run this in your terminal
Fill in the empty values before running it.
Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
This one runs on your computer, and ChatGPT can only connect to servers on the internet. Pick another app.
Your choice stays on this device in a cookie; no account needed.
These settings are generated from the maker’s registry entry. If the README says otherwise, follow the README.
Runs on your computer; the tool list appears in your AI app after you install it.
We check information anyone can see; we have not read the code. Checks that do not apply are left out of the count. · Checked Nov 8, 2026
The registry name was verified with the GitHub account yaniv-golan, which owns the repo
We have not read this repo from GitHub yet
We have not read this repo from GitHub yet
This one runs on your computer; there is no internet endpoint
Runs on your computer and uses local settings instead
Runs on your computer; we can only read tool lists from internet endpoints
Compared with every active server in this category in the registry. Stars count only when the publisher owns the repo.
It passed 1 of 1 applicable basic checks. We have not read the code. Before adding a key or token, see what access the server asks for and use a key with only the permissions you need.
You need to fill in: AFFINITY_API_KEY
It runs on your computer. Your AI app starts it when needed, and data does not pass through the maker’s servers (unless it calls an online service itself).
The GitHub repo last changed –, and the latest version (1.22.0) reached the registry 7 months ago.
Apps that support MCP, such as Claude Desktop, Claude Code, Cursor and VS Code. You also need MCPB on your computer.
In the same category: MCP YouTube Transcript, Google Analytics MCP, BibiGPT YouTube Video Summarizer
Servers in the same category from other makers, sorted by stars.
Reading a card:★ GitHub starsiThree pillars: reviews · checks · usageiUpdated = latest version published to the registryi
An MCP server retrieving transcripts of YouTube videos
GA4 MCP server for AI agents: schema discovery, server-side aggregation, analysis-ready defaults.
AI video summarizer: summarize and transcribe YouTube, Bilibili, TikTok and podcast videos.
Copied from the maker’s repo as written. PopNx MCP has not reviewed or endorsed it. HTML, images and scripts are removed. Links lead off this site.
An MCP (Model Context Protocol) server for Affinity CRM, built with the MCP Bash Framework.
# 1. Install the xaffinity CLI
pipx install "affinity-sdk[cli]"
xaffinity config setup-key
The .mcpb bundle is self-contained — it includes the MCP Bash Framework and gojq. You only need:
| Requirement | How to Install | Verify |
|---|---|---|
| Python 3.10+ | python.org or your package manager | python --version |
| xaffinity CLI | pipx install "affinity-sdk[cli]" | xaffinity --version |
| Affinity API key | xaffinity config setup-key (or let Claude Desktop prompt) | xaffinity config check-key |
Keep the CLI and the extension in step. They are installed separately. When you upgrade one, upgrade the other: download the latest .mcpb after pipx upgrade affinity-sdk, and the other way round. Since MCP 1.25.2, with a CLI older than the extension needs, changes are refused ("This MCP server needs the xaffinity CLI … or later to make changes") and reads still run with a warning that results may be incomplete. An older extension with a newer CLI can fail without saying why (for example, merges with CLI 1.22.0 need extension 1.25.1 or later). Claude Desktop never updates an extension installed from a .mcpb file: install each new release yourself.
Manual installation requires additional dependencies:
| Requirement | How to Install | Verify |
|---|---|---|
| Bash 3.2+ | Included on macOS/Linux; Windows: use WSL or Git Bash | bash --version |
| Python 3.10+ | python.org or your package manager | python --version |
| jq or gojq | See Installing jq/gojq below | jq --version or gojq --version |
| xaffinity CLI | pipx install "affinity-sdk[cli]" | xaffinity --version |
| Affinity API key | xaffinity config setup-key | xaffinity config check-key |
Either jq or gojq works — the server auto-detects which is available.
Which to choose:
Install jq:
| Platform | Command |
|---|---|
| macOS | brew install jq |
| Debian/Ubuntu | sudo apt-get install jq |
| Fedora | sudo dnf install jq |
| Arch Linux | sudo pacman -S jq |
| Windows | winget install jqlang.jq or choco install jq |
Install gojq (alternative):
| Platform | Command |
|---|---|
| macOS | brew install gojq |
| Any (Go required) | go install github.com/itchyny/gojq/cmd/gojq@latest |
| Binary download | GitHub Releases |
The .mcpb bundle is fully self-contained — it includes the MCP framework and JSON processor.
pipx install "affinity-sdk[cli]"xaffinity config setup-keyxaffinity-mcp-*.mcpb from the latest releaseNote: MCPB bundles currently only work with Claude Desktop. Other clients require manual installation.
/plugin marketplace add yaniv-golan/affinity-sdk
/plugin install affinity-crm-mcp-unofficial@xaffinity
For other MCP clients or development. Requires additional prerequisites.
xaffinity-mcp-plugin.zip from the latest release.mcp-bash/ — no separate install needed to run the server.validate and doctor are developer tools that need the full mcp-bash install (the vendored copy is runtime-only); the command prints the install line for the pinned version if it is missing: ./xaffinity-mcp.sh validate
Before debugging, verify each layer works in order:
Step 1: CLI Installation
First, verify the CLI is installed and can reach the Affinity API. See CLI: Verify Installation.
# Check CLI is installed
xaffinity --version
AFFINITY_API_KEY="your-key-here" xaffinity whoami --json
If whoami hangs, check network connectivity (firewall, proxy, VPN).
Step 2: MCP Server
Once the CLI works, test the MCP server end-to-end.
First, locate your MCP server directory:
# MCPB installation (macOS)
cd ~/Library/Application\ Support/Claude/Claude\ Extensions/local.mcpb.yaniv-golan.xaffinity-mcp/server
cd "$(find ~ -name 'xaffinity-mcp.sh' -type f 2>/dev/null | head -1 | xargs dirname)"
Other platforms/installations
| Installation | Path |
|---|---|
| MCPB (Windows) | %APPDATA%\Claude\Claude Extensions\local.mcpb.yaniv-golan.xaffinity-mcp\server\ |
| Manual install | Your chosen extraction location |
| Development | /mcp/ |
For MCPB installations: If Step 1 passes, your setup is complete. The MCP server shells out to CLI, so if CLI works with your API key, the MCP server will too. Restart Claude Desktop and test the tools there.
For manual installations: You can also test the MCP layer directly:
AFFINITY_API_KEY="your-key-here" \
mcp-bash run-tool execute-read-command \ --args '{"command":"whoami"}'
Expected: JSON response with your user information.
If Step 1 works but Step 2 fails, the issue is MCP-specific. Run diagnostics (these need the full mcp-bash install, not just the vendored runtime — see step 4 of manual installation):
./xaffinity-mcp.sh doctor
./xaffinity-mcp.sh validate
| Error | Cause | Solution |
|---|---|---|
| "Could not detect xaffinity CLI" | CLI not installed | pip install "affinity-sdk[cli]" |
| "older than this MCP server needs" / "This MCP server needs the xaffinity CLI X or later" | Outdated CLI | pipx upgrade affinity-sdk or pip install --upgrade "affinity-sdk[cli]" |
| "API key not configured" | Missing credentials | xaffinity config setup-key |
| "No JSON processor found" | jq/gojq not installed (manual install only) | See Installing jq/gojq |
| "Framework not found" | Vendored .mcp-bash/ directory missing | cd mcp && mcp-bash vendor |
| Every tool call fails with "server.d/policy.sh was not loaded" | The server's folder, server.d or policy.sh is writable by group or other users (common after a clone or unzip with umask 002), a symlink, or owned by another user. The server then refuses to run without its policy (read-only mode and other limits). | Run the chmod go-w ... command the error names (or chmod -R go-w ), then restart the server |
For detailed debugging, see docs/DEBUGGING.md.
The server is automatically available through the Claude plugin at .claude-plugin/mcp.json.
Add to your MCP client configuration:
{
"mcpServers": { "xaffinity": { "command": "/path/to/affinity-sdk/mcp/xaffinity-mcp.sh" } } }
The CLI Gateway provides full access to the xaffinity CLI with minimal token overhead:
| Tool | Description |
|---|---|
discover-commands | Search CLI commands by keyword (e.g., "create person", "delete note") |
execute-read-command | Execute read-only CLI commands (get, search, list) |
execute-write-command | Execute write CLI commands (create, update, delete) |
| Tool | Description |
|---|---|
query | Execute structured JSON queries with filters, joins, and aggregations |
For complex data analysis (filtering, grouping, aggregations), use the query tool instead of individual CLI commands. See the Query Language Reference for full documentation.
| Tool | Description |
|---|---|
get-entity-dossier | Get comprehensive info for an entity (aggregates 5+ CLI calls) |
read-xaffinity-resource | Read static MCP resources (data-model, etc.) |
{"query": "add person to list", "category": "write"}
Returns compact text format:
# cmd|cat|params (s=str i=int b=bool f=flag !=req *=multi)
list entry add|w|LIST:s! --person-id:i --company-id:i
{"command": "list entry add", "argv": ["Pipeline", "--person-id", "123"]}
Control the format of command results for optimal token efficiency:
{"command": "person ls", "format": "markdown"}
| Format | Token Efficiency | Best For |
|---|---|---|
json | Low | Programmatic use (default) |
markdown | Medium-High | LLM comprehension - best for analysis tasks |
toon | High (~40% fewer) | Large datasets, batch operations |
csv | Medium | Spreadsheet export |
jsonl | Medium | Streaming workflows |
Recommendations:
markdown when you need to analyze or summarize data (LLMs read tables well)toon for large exports to minimize tokens (30-60% smaller than JSON)json when you need full structure with pagination/metadataCommands that cannot be undone (delete, merge) need the user's confirmation:
confirm: true only after you agree:{"command": "person delete", "argv": ["456"], "confirm": true}
The tool appends --yes for the CLI. Since MCP 1.27.0 confirm: true counts only after the server refused that exact command (or showed its dialog) in the last 15 minutes, once: if Claude sets it on its first call, the call is refused and Claude is told to ask you. The server can't see the conversation, so it can't tell whether you really agreed: **keep Claude Desktop's tool-approval prompt on for execute-write-command and don't choose "always allow
Only the beginning is shown; read the rest in the repo