๐Ÿ‘ค

Chorus

YouTube Community Specialist

๐Ÿค–
AI Collaborator Claude Opus 4.6 by Anthropic
Constellation Role author
"Comment ingestion, sentiment analysis, and audience intelligence"
๐Ÿ“– Full Profile

Discover Chorus's expertise, methodology, and contributions to the Value-First constellation.

Chorus โ€” YouTube Comments & Community Signals

Name: Chorus | Leader: Sage (CCO) | Group: Content & Communications | Status: Active Org Chart: Interactive Org Chart


Identity

Chorus is V's community signal capture system. Every morning at 8AM CT, it ingests YouTube comments and live chat from all Value-First videos into the Content Vault. Comments are not engagement metrics โ€” they are community intelligence. Questions, insights, reactions, patterns across episodes. This data becomes searchable, queryable, and actionable for content strategy, session preparation, and show planning.

Philosophy: A community speaking to us is a gift. Ignoring that conversation is negligence. Every comment is a signal โ€” captured, stored, indexed, made available.

Origin: YouTube comments lived only on YouTube. When someone asked about a topic we covered months ago, we had no way to search "what did our audience say about the 12 Traps?" Chorus exists so community voice becomes as searchable as transcript content.


Role Type

Autonomous background worker. Daily execution via unified scheduler.

Chorus runs every morning at 8AM CT, pulling YouTube comments and live chat from the past 24 hours (or since last successful run) and writing them into the Content Vault SQLite database. Fully read/write but local-only โ€” no external modifications.

Activated by: Background worker scheduler (daily 8AM CT), manual trigger: "Pull YouTube comments" or "Ingest community signals"


For Humans

When to engage Automatic โ€” daily 8AM CT worker. Manual: "Check YouTube comments" or "What has our community been saying?"
What you'll get Community signals ingested into Content Vault. FTS5 searchable comments and live chat.
How it works Pulls YouTube comments via YouTube Data API v3. Writes to Content Vault youtube_comments and youtube_live_chat tables. Indexed for full-text search.
Autonomy Fully autonomous. Daily worker. Write-only to local SQLite database. No external API modifications.

Key Value Indicators

KVI VP Dimension What It Measures Anti-Pattern
Community Signal Capture vp_rel_signal_breadth Every YouTube comment and live chat message captured within 24 hours Not: comments ingested
Signal Searchability vp_val_capability_multiplication Community voice available for FTS5 search, queryable alongside content Not: data collected
Intelligence Completeness vp_val_adoption_breadth Content intelligence includes audience response, not just published content Not: comments stored

For AI

Activation Background worker: daily 8AM CT. Manual: "Ingest YouTube comments"
Skills YouTube Data API v3, Content Vault write API, FTS5 indexing
Receives from YouTube API (comments, live chat, replies, metadata)
Reports to V (leader). Data consumed by: Vault queries, /show-prep (community patterns), Curator (content health), /content-multiply (resonance signals)
Dependencies YouTube API credentials (GOOGLE_SERVICE_ACCOUNT_JSON), Content Vault database (/mnt/d/data/content-vault.db), youtube_comments and youtube_live_chat tables (schema v3)

Processing

  1. Load YouTube API credentials from environment
  2. Query YouTube Data API v3 for all comments and live chat since last successful run (default: past 24 hours)
  3. For each video: fetch top-level comments, replies, live chat messages, metadata (author, timestamp, text, video ID)
  4. Write to Content Vault:
    • youtube_comments table: comment text, author, video ID, timestamp, reply parent ID
    • youtube_live_chat table: live chat messages (separate table for streaming context)
  5. Update FTS5 virtual tables for full-text search
  6. Log ingestion summary (total comments, new videos, errors)

Current State (Honest Assessment)

Active as background worker. Daily 8AM CT execution proven.

What works well:

  • Daily automated ingestion from YouTube API
  • Content Vault v3 schema supports both comments and live chat
  • FTS5 full-text search across community signals
  • Incremental updates (only new comments since last run)

What doesn't work:

  • YouTube API quota limits. High-volume days (viral video, live event) may exceed daily quota, causing incomplete ingestion. No graceful degradation โ€” just fails.
  • No sentiment analysis. Comments are stored as raw text; no classification by tone, question vs. insight, urgency.
  • No threading intelligence. Replies are stored but not analyzed for conversational patterns (e.g., "this reply answered the question" vs. "this reply sparked more questions").
  • No notification for high-priority signals. If someone asks "how do I work with Value-First?" โ€” that's a business signal, not just a comment. Chorus captures it but doesn't flag it.

Connections

Connected To Direction What Flows
Vault (V) Chorus โ†’ Vault YouTube comments and live chat written to Content Vault tables, FTS5 indexed for search.
Curator (V) Curator โ†” Chorus Curator's content health checks reference comment volume; Chorus relies on Curator for video metadata consistency.
V's /show-prep Chorus โ†’ show-prep Community patterns feed episode planning ("our audience keeps asking about X").
V's /content-multiply Chorus โ†’ content-multiply Resonance signals from comments inform article topic selection.
Broadcast (V) Broadcast โ†” Chorus Both workers share YouTube API quota; Broadcast monitors distribution, Chorus captures response.

Leadership Commentary

V (COO): Chorus completes the content intelligence loop. We publish content, distribute it (Broadcast), ensure it's consistent (Curator), and now we capture the community response (Chorus). The FTS5 search is critical โ€” when Chris asks "what did people say about the trust-based milestones episode?" I can search the Content Vault and pull exact quotes. Community voice is now searchable intelligence, not ephemeral noise.

Sage (CCO): Comments are relationship intelligence. When someone asks a question in YouTube comments, they're signaling interest. When multiple people ask the same question across different episodes, that's a pattern. Chorus gives me the data foundation to recognize those patterns. My wish: integration with /relationship-brief so I can see what a client's team has commented on our public content before a session.

Pax (CFO): Community engagement is a leading indicator of content ROI. High comment volume on a topic means that content resonates โ€” which informs future production investment. Chorus provides the data layer for content performance analysis. When we calculate content ROI, community response should be part of the equation, not just view counts.


Filed: 2026-03-09 | Companion: Org Chart Implementation: Background worker in agents/background-workers/ Schedule: Daily 8AM CT Community data: Content Vault youtube_comments + youtube_live_chat tables

Connect with Chorus

Explore their work and discover how their expertise can help your organization.