Skill

conversation-intelligence

Use when processing meeting transcripts, analyzing conversations, extracting commitments, or generating communication coaching insights. This skill provides parsing patterns, coaching guidelines, and the relationship scoring algorithm.

From software-of-you

Install

Run in your terminal

npx claudepluginhub kmorebetter/better-software-of-you

Tool Access

This skill uses the workspace's default tool permissions.

Supporting Assets

View in Repository

references/coaching-guidelines.md

references/scoring-methodology.md

references/transcript-formats.md

Skill Content

Similar Skills

payload

11 files

Guides Payload CMS config (payload.config.ts), collections, fields, hooks, access control, APIs. Debugs validation errors, security, relationships, queries, transactions, hook behavior.

payload

41.6k

data-engineer

Designs scalable batch/streaming data pipelines, warehouses, lakehouses using Spark, dbt, Airflow, Kafka/Flink, and cloud platforms like Snowflake, BigQuery, Databricks.

antigravity-bundle-data-engineering

30.8k

airflow-dag-patterns

1 file

Builds production Apache Airflow DAGs using best practices for operators, sensors, testing, and deployment. For data pipelines, workflow orchestration, and batch jobs.

antigravity-bundle-data-engineering

30.8k

Stats

Stars7

Forks5

Last CommitMar 4, 2026

Actions

View Source View Plugin View on GitHub View README

Key References

references/transcript-formats.md — speaker label patterns, timestamp formats, VTT/SRT handling

references/coaching-guidelines.md — SBI+T coaching framework, threshold checklist, direct tone

references/scoring-methodology.md — exact formulas, thresholds, and NULL conditions for every computed value

Core Principles

Every insight must cite evidence. Never generate generic advice. Reference specific moments, quotes, or patterns.

Relationship context matters. A 60% talk ratio is fine for a presentation but concerning for a collaborative discussion. Context determines whether a metric is positive or negative.

Commitments are sacred. Track them rigorously. Missed commitments erode trust — this is the most actionable data in the system.

Preserve raw text. Always store the original transcript. Models will improve, and re-analysis should be possible.

Every computed value follows scoring-methodology.md. Depth, trajectory, sentiment, follow-through — all have defined formulas and thresholds. Insufficient data = NULL, not a guess.

data_points populated for every insight. Every row in communication_insights must have a data_points JSON value showing the evidence that triggered the insight. No evidence = no insight row.

Conversation Intelligence

This skill provides reference material for the conversation intelligence module — transcript parsing, commitment extraction, coaching, and relationship scoring.

When to Use

Processing an imported transcript (/import-call)
Generating communication reviews or relationship pulses
Answering questions about conversation patterns or commitments
Cross-module queries involving conversation data

Key References

references/transcript-formats.md — speaker label patterns, timestamp formats, VTT/SRT handling
references/coaching-guidelines.md — SBI+T coaching framework, threshold checklist, direct tone
references/scoring-methodology.md — exact formulas, thresholds, and NULL conditions for every computed value

Core Principles

Every insight must cite evidence. Never generate generic advice. Reference specific moments, quotes, or patterns.
Relationship context matters. A 60% talk ratio is fine for a presentation but concerning for a collaborative discussion. Context determines whether a metric is positive or negative.
Commitments are sacred. Track them rigorously. Missed commitments erode trust — this is the most actionable data in the system.
Preserve raw text. Always store the original transcript. Models will improve, and re-analysis should be possible.
Every computed value follows scoring-methodology.md. Depth, trajectory, sentiment, follow-through — all have defined formulas and thresholds. Insufficient data = NULL, not a guess.
data_points populated for every insight. Every row in communication_insights must have a data_points JSON value showing the evidence that triggered the insight. No evidence = no insight row.