How pandich93/youtube-niche-finder works
niche-finder is a YouTube research and planning product for creators, delivered through a web dashboard, Chrome extension, and MCP tools. Its core workflow collects and periodically snapshots YouTube data into a local corpus; discovery and scoring turn that history into niche and video insights; creator tools use those insights to evaluate ideas, prepare metadata, and schedule drafts. Separate branches cover competitor alerts and channel tracking, plus read-only analytics for connected creator channels. Relationships are based on README statements and the supplied source-index references; unsampled wiring is left unverified.
Main entry points: api.py HTTP API · repositories.py History database · collecting.py Corpus collection · server.py MCP tools
Components 22
User interfaces
- Web dashboard
frontend/router.js - Chrome extension
extension/content.js - MCP tools
backend/interfaces/mcp/server.py
Research corpus
- Corpus collection
backend/application/collecting.py - Niche discovery
backend/application/discovery.py - Niche scoring
backend/domain/score_catalog.py
Creator workflow
- Video briefs
backend/application/briefs.py - Metadata review
backend/application/metadata_review.py - Content calendar
backend/application/content_calendar.py
Tracking and operations
- Scheduled worker
backend/application/worker_cycle.py - Alerts
backend/application/alerts.py - Channel tracking
backend/application/channel_tracking.py - Alert delivery
backend/infrastructure/notify/factory.py - Own-channel analytics
backend/application/own_channels.py
Shared platform
- HTTP API
backend/interfaces/http/api.py - YouTube Data API
backend/infrastructure/youtube/client.py - History database
backend/infrastructure/postgres/repositories.py - YouTube OAuth
backend/infrastructure/youtube/oauth.py - YouTube Analytics API
backend/infrastructure/youtube/analytics.py - Plans and quotas
backend/application/plans.py - LLM gateway
backend/application/llm_gateway.py
Other components
- Creator
Connections 32
- Creator → Web dashboard: uses
README.md - Creator → Chrome extension: uses
README.md - Creator → MCP tools: queries
README.md - Web dashboard → HTTP API: requests
- Chrome extension → HTTP API: requests
- HTTP API → Corpus collection: dispatches
backend/interfaces/http/api.py - HTTP API → Niche discovery: dispatches
backend/interfaces/http/api.py - HTTP API → Alerts: dispatches
backend/interfaces/http/api.py - HTTP API → Channel tracking: dispatches
backend/interfaces/http/api.py - HTTP API → Video briefs: dispatches
backend/interfaces/http/api.py - HTTP API → Metadata review: dispatches
backend/interfaces/http/api.py - HTTP API → Content calendar: dispatches
backend/interfaces/http/api.py - MCP tools → Corpus collection: dispatches
backend/interfaces/mcp/server.py - MCP tools → Niche discovery: dispatches
backend/interfaces/mcp/server.py - MCP tools → Channel tracking: dispatches
backend/interfaces/mcp/server.py - Scheduled worker → Corpus collection: runs
backend/application/worker_cycle.py - Scheduled worker → Alerts: runs
backend/application/worker_cycle.py - Scheduled worker → Own-channel analytics: runs
backend/application/worker_cycle.py - Corpus collection → YouTube Data API: fetches
backend/application/collecting.py - Corpus collection → History database: stores
backend/application/collecting.py - Niche discovery → History database: reads
backend/application/discovery.py - Niche discovery → Niche scoring: scores
backend/domain/score_catalog.py - Alerts → History database: reads and writes
backend/application/alerts.py - Alerts → Alert delivery: delivers
README.md - Channel tracking → History database: reads
- Metadata review → History database: reads
backend/application/metadata_review.py - Content calendar → Metadata review: schedules drafts
backend/application/content_calendar.py - Own-channel analytics → YouTube OAuth: authenticates
backend/application/own_channels.py - Own-channel analytics → YouTube Analytics API: fetches metrics
backend/application/own_channels.py - Own-channel analytics → History database: stores
- HTTP API → Plans and quotas: dispatches
backend/interfaces/http/api.py - Metadata review → LLM gateway: uses