Keyapi Youtube Channel Analysis

Discover, profile, and analyze YouTube channels — retrieve channel metadata, video libraries, convert between channel IDs and URLs, search channels by keywor...

Security Scan

The skill's code, required environment variables, and runtime instructions are coherent with a KeyAPI-based YouTube channel analysis tool; the only notable behaviors are local caching and writing the provided API token to a .env file in the skill directory.

Purpose & Capability

The name/description promise YouTube channel discovery and analysis via the KeyAPI MCP service. The skill requires NODE and a KEYAPI_TOKEN and includes a small runner (scripts/run.js) plus a dependency on @modelcontextprotocol/sdk — all expected for calling the KeyAPI MCP. There are no unrelated credentials or binaries requested.

Instruction Scope

SKILL.md directs the agent to install dependencies (npm install), set KEYAPI_TOKEN, and run node scripts/run.js to list or call MCP tools. The code reads a .env file, will prompt for a token if missing, and will persist the token to a .env file when entered. It also writes cache files (.keyapi-cache) and can write results to an --output path. These file I/O actions are within the stated purpose but are persistent and worth being aware of.

Install Mechanism

No remote download/install script is included; the package.json declares a single SDK dependency and the SKILL.md asks the user to run npm install. This is a standard, proportional install approach (no arbitrary external archives or URL downloads).

Credentials

Only KEYAPI_TOKEN (primary credential) and an optional KEYAPI_SERVER_URL override are used. That aligns with a service-client skill. The code does persist the token to a local .env file (unencrypted), which the user should avoid committing to source control.

Persistence & Privilege

The skill is not forced-always or otherwise privileged. It persists its own cache (.keyapi-cache) and may create a .env file in the skill directory; it does not modify other skills or global agent configuration. Autonomous invocation is allowed by default (platform normal), but this skill's footprint is limited to its directory by default.

Assessment

This skill appears to do what it says: it calls KeyAPI MCP tools to analyze YouTube channels and needs only KEYAPI_TOKEN and node. Before installing: 1) Do not commit the .env file that the tool may create — it stores your token in plaintext in the skill directory. 2) Prefer creating and using a least-privilege token from keyapi.ai and rotate it if exposed. 3) Review or run the tool in an isolated environment (or container) if you want to limit where it can write cache or output files. 4) If you plan to override KEYAPI_SERVER_URL, verify the server is trustworthy — that override lets the runner talk to any endpoint. If you want, inspect the remainder of scripts/run.js (the network-call portion) to confirm no unexpected remote hosts are contacted beyond the configured MCP server.

Discover, profile, and analyze YouTube channels — from identity resolution and metadata retrieval to video inventory, channel search, and advanced filtered discovery.

This skill provides comprehensive YouTube channel intelligence using the KeyAPI MCP service. It enables detailed channel profile retrieval (including a two-step process for full metadata), video library enumeration, bidirectional conversion between channel IDs and URLs, keyword-based channel search, within-channel content search, advanced filtered search across YouTube, and search suggestion retrieval — all through a cache-first workflow.

Use this skill when you need to:

• Retrieve full channel metadata including description, subscriber count, view count, join date, and social links
• Convert between channel URLs (various formats) and channel IDs, or resolve @username handles
• Enumerate a channel's published video library with pagination support
• Search for channels by keyword with result filtering to channel-type only
• Search within a specific channel's content by keyword
• Perform advanced YouTube searches with filters (upload time, duration, content type, features, sort order)
• Get real-time search suggestions for keyword research and autocomplete

author: KeyAPI license: MIT repository: https://github.com/EchoSell/keyapi-skills

Prerequisites

Requirement	Details
KEYAPI_TOKEN	A valid API token from keyapi.ai. If you don't have one, register at the site to obtain your free token. Set it as an environment variable: export KEYAPI_TOKEN=your_token_here
Node.js	v18 or higher
Dependencies	Run npm install in the skill directory to install @modelcontextprotocol/sdk

author: KeyAPI license: MIT repository: https://github.com/EchoSell/keyapi-skills

MCP Server Configuration

All tool calls in this skill target the KeyAPI YouTube MCP server:

Server URL : https://mcp.keyapi.ai/youtube/mcp
Auth Header: Authorization: Bearer $KEYAPI_TOKEN

Setup (one-time):

# 1. Install dependencies
npm install

# 2. Set your API token (get one free at https://keyapi.ai/)
export KEYAPI_TOKEN=your_token_here

# 3. List all available tools to verify the connection
node scripts/run.js --platform youtube --list-tools

author: KeyAPI license: MIT repository: https://github.com/EchoSell/keyapi-skills

Analysis Scenarios

Channel Identity & Metadata Nodes

User Need	Node(s)	Best For
Full channel metadata (two-step process)	get_channel_description	Complete channel audit — call twice: first with channel_id, then with continuation_token from response
Extract channel ID from any URL format	get_channel_id_from_url	URL normalization — supports @username, /channel/, /c/, /user/ formats
Get @username handle from channel ID	get_channel_url_from_channel_id	Reverse lookup — channel ID → @handle
Resolve channel name to channel ID	get_channel_id	Keyword-based ID lookup by channel name

Channel Content Nodes

User Need	Node(s)	Best For
List all videos from a channel	get_channel_videos	Video inventory, content cadence analysis — requires UC-format channel ID

Search & Discovery Nodes

User Need	Node(s)	Best For
Search for channels by keyword	search_channels	Channel discovery — returns only channel-type results, filters out videos/playlists
Search within a specific channel	search_channel	In-channel content search by keyword — requires channel ID
Advanced filtered search across YouTube	general_search_with_filters	Precision discovery — filter by upload time, duration, content type, features, sort order
Get search suggestions / autocomplete	get_search_suggestions	Keyword research, query expansion — returns 10–20 suggestions in <1 second

author: KeyAPI license: MIT repository: https://github.com/EchoSell/keyapi-skills

Workflow

Step 1 — Identify the Channel Analysis Objective and Select Nodes

Clarify the research goal and map it to one or more nodes. Common patterns:

• Channel profile deep-dive: Use get_channel_description twice — first call with channel_id returns basic info + continuation_token; second call with that token returns advanced info (join date, social links, view count).
• URL → ID conversion: Use get_channel_id_from_url to normalize any channel URL format to a channel ID.
• ID → @handle conversion: Use get_channel_url_from_channel_id to get the user-friendly @username from a channel ID.
• Channel discovery: Use search_channels with keyword → optionally deepen with get_channel_description for each result.
• Video inventory: Use get_channel_videos with channel_id (must be UC-format) → paginate with continuation_token.
• Advanced search: Use general_search_with_filters with upload time, duration, content type, and sort filters.

Channel ID formats

• UC-format: 24-character string starting with UC (e.g., UCeu6U67OzJhV1KwBansH3Dg) — required for get_channel_videos and get_channel_description.
• @username format: User-friendly handle (e.g., @CozyCraftYT) — use get_channel_id_from_url to convert to UC-format.
• URL formats: Supports youtube.com/@username, youtube.com/channel/UCxxxxxx, youtube.com/c/name, youtube.com/user/name.

get_channel_description two-step process

This endpoint requires two sequential calls to retrieve complete channel information:

1. First call: Pass channel_id → returns basic info (name, description, subscriber count, video count, banner) + continuation_token.
2. Second call: Pass continuation_token from step 1 → returns advanced info (join date, social media links, featured sections, total view count).

Do NOT call both simultaneously — the second call depends on the token from the first.

Step 2 — Retrieve API Schema

Before calling any node, inspect its input schema to confirm required parameters and available options:

node scripts/run.js --platform youtube --schema <tool_name>

# Examples
node scripts/run.js --platform youtube --schema get_channel_description
node scripts/run.js --platform youtube --schema general_search_with_filters

Step 3 — Call APIs and Cache Results Locally

Execute tool calls and persist responses to the local cache.

Calling a tool:

node scripts/run.js --platform youtube --tool <tool_name> \
  --params '<json_args>' --pretty

# Skip cache for fresh results
node scripts/run.js --platform youtube --tool <tool_name> \
  --params '<json_args>' --no-cache --pretty

Example — get full channel metadata (two-step):

# Step 1: Get basic info + continuation_token
node scripts/run.js --platform youtube --tool get_channel_description \
  --params '{"channel_id":"UCeu6U67OzJhV1KwBansH3Dg","language_code":"en-US","country_code":"US","need_format":true}' --pretty

# Step 2: Get advanced info using continuation_token from Step 1
node scripts/run.js --platform youtube --tool get_channel_description \
  --params '{"continuation_token":"<token_from_step_1>","language_code":"en-US","country_code":"US","need_format":true}' --pretty

Example — convert channel URL to ID:

node scripts/run.js --platform youtube --tool get_channel_id_from_url \
  --params '{"channel_url":"https://www.youtube.com/@CozyCraftYT"}' --pretty

Example — get @handle from channel ID:

node scripts/run.js --platform youtube --tool get_channel_url_from_channel_id \
  --params '{"channel_id":"UCeu6U67OzJhV1KwBansH3Dg"}' --pretty

Example — get channel videos (first page):

node scripts/run.js --platform youtube --tool get_channel_videos \
  --params '{"channel_id":"UCJHBJ7F-nAIlMGolm0Hu4vg","language_code":"en-US","country_code":"US"}' --pretty

Example — paginate to next video page:

node scripts/run.js --platform youtube --tool get_channel_videos \
  --params '{"channel_id":"UCJHBJ7F-nAIlMGolm0Hu4vg","continuation_token":"<token>"}' --pretty

Example — search channels by keyword:

# First page
node scripts/run.js --platform youtube --tool search_channels \
  --params '{"keyword":"fitness","need_format":true}' --pretty

# Next page
node scripts/run.js --platform youtube --tool search_channels \
  --params '{"continuation_token":"<token>","need_format":true}' --pretty

Example — search within a channel:

node scripts/run.js --platform youtube --tool search_channel \
  --params '{"channel_id":"UCXuqSBlHAE6Xw-yeJA0Tunw","search_query":"AMD","language_code":"en","country_code":"US"}' --pretty

Example — advanced filtered search:

node scripts/run.js --platform youtube --tool general_search_with_filters \
  --params '{"search_query":"Python tutorial","upload_time":"month","duration":"medium","content_type":"video","feature":"hd","sort_by":"view_count","language_code":"en-US","country_code":"US"}' --pretty

Example — get search suggestions:

node scripts/run.js --platform youtube --tool get_search_suggestions \
  --params '{"keyword":"machine learning","language":"en","region":"US"}' --pretty

Example — resolve channel name to ID:

node scripts/run.js --platform youtube --tool get_channel_id \
  --params '{"channel_name":"LinusTechTips"}' --pretty

Pagination:

Endpoint	Pagination	Notes
get_channel_description	continuation_token	Two-step: first call returns token for second call
get_channel_videos	continuation_token	From previous response
search_channels, search_channel, general_search_with_filters	continuation_token	From previous response
get_channel_id_from_url, get_channel_url_from_channel_id, get_channel_id, get_search_suggestions	—	Single-call; no pagination

Cache directory structure:

.keyapi-cache/
└── YYYY-MM-DD/
    ├── get_channel_description/
    │   └── {params_hash}.json
    ├── get_channel_id_from_url/
    │   └── {params_hash}.json
    ├── get_channel_url_from_channel_id/
    │   └── {params_hash}.json
    ├── get_channel_videos/
    │   └── {params_hash}.json
    ├── get_channel_id/
    │   └── {params_hash}.json
    ├── search_channels/
    │   └── {params_hash}.json
    ├── search_channel/
    │   └── {params_hash}.json
    ├── general_search_with_filters/
    │   └── {params_hash}.json
    └── get_search_suggestions/
        └── {params_hash}.json

Cache-first policy:

Before every API call, check whether a cached result already exists. If valid, load from disk and skip the API call.

Step 4 — Synthesize and Report Findings

For channel analysis:

1. Channel Overview — Name, @handle, channel ID, subscriber count, total video count, total view count, join date, verification status.
2. Channel Description — Bio, featured links, social media connections, country/language.
3. Content Inventory — Video count, upload frequency, most recent video date, content categories.
4. Audience Signals — Subscriber-to-view ratio, engagement indicators from recent videos.
5. Featured Sections — Playlists, channels, or content highlighted by the creator.

For channel discovery:

1. Search Results Overview — Matched channels, subscriber counts, verification status, content focus.
2. Comparative Analysis — Side-by-side metrics for shortlisted channels.
3. Keyword Expansion — Search suggestions for related queries, trending topics.

author: KeyAPI license: MIT repository: https://github.com/EchoSell/keyapi-skills

Common Rules

Rule	Detail
Channel ID format	get_channel_videos and get_channel_description require UC-format channel IDs (24 characters starting with UC). Use get_channel_id_from_url to convert @username or other URL formats.
get_channel_description two-step	Critical: This endpoint requires two sequential calls. First call with channel_id returns basic info + continuation_token. Second call with that token returns advanced info. Do NOT call both simultaneously.
URL format support	get_channel_id_from_url supports multiple formats: @username, /channel/UCxxxxxx, /c/name, /user/name.
search_channels vs search_channel	search_channels searches across YouTube for channels by keyword. search_channel searches within a specific channel's content.
need_format parameter quirk	The get_channel_videos schema has a trailing space in the parameter name: "need_format " (with space). However, the API accepts both forms. Omit this parameter in examples to avoid confusion. Available on get_channel_description and search_channels without the space issue.
Language / region params	language_code and country_code influence result language and regional content bias. Use en-US / US for English-language global results.
Success check	code = 0 → success. Any other value → failure. Always check the response code before processing data.
Retry on 500	If code = 500, retry the identical request up to 3 times with a 2–3 second pause between attempts before reporting the error.
Cache first	Always check the local .keyapi-cache/ directory before issuing a live API call.

author: KeyAPI license: MIT repository: https://github.com/EchoSell/keyapi-skills

Error Handling

Code	Meaning	Action
0	Success	Continue workflow normally
400	Bad request — invalid or missing parameters	Validate input against the tool schema; check channel ID format (must be UC-format for some endpoints), URL format, and required fields
401	Unauthorized — token missing or expired	Confirm KEYAPI_TOKEN is set correctly; visit keyapi.ai to renew
403	Forbidden — plan quota exceeded or feature restricted	Review plan limits at keyapi.ai
404	Resource not found — channel may not exist or URL is invalid	Verify the channel ID or URL; the channel may have been deleted or made private
429	Rate limit exceeded	Wait 60 seconds, then retry
500	Internal server error	Retry up to 3 times with a 2–3 second pause; if it persists, log the full request and response and skip this node
Other non-0	Unexpected error	Log the full response body and surface the error message to the user