Add Conversation Insights Coach ability#247
Conversation
Passive background daemon that analyzes HOW the user communicates: - Tracks filler words (um, like, you know, basically, etc.) - Detects hedging language (maybe, I think, sort of, etc.) - Measures vocabulary diversity and unique word usage - Calculates question-to-statement ratio - Monitors average utterance length and speech patterns Features: - On-demand spoken coaching reports with trends and milestones - Real-time nudge system (optional) for filler word awareness - Customizable filler word goals with progress tracking - 30-day history with daily comparisons and streak tracking - Milestone celebrations (personal bests, flawless sessions) - Detail drill-downs for specific communication areas - Session persistence across reconnects All communication analysis is passive and happens silently every 20 seconds. Reports are generated using LLM-based coaching prompts for natural, warm delivery.
🔀 Branch Merge CheckPR direction: ✅ Passed — |
✅ Ability Validation Passed |
✅ Community PR Path Check — PassedAll changed files are inside the |
🔍 Lint Results✅
|
- Empty __init__.py (no comments or code) - Remove unnecessary f-string prefix on line 447
uzair401
left a comment
There was a problem hiding this comment.
Approved for the community.
We will proceed with adopting this ability for a 30-day evaluation period to monitor performance, identify any bugs, and implement any necessary fixes or optimizations. If the ability proves stable and meets our quality standards during this period, we will move forward with publishing it to the marketplace, ensuring full credit is given to the original author for their work and contribution.
We truly appreciate your effort and encourage you to continue developing and submitting new abilities. Contributions like yours help strengthen the ecosystem, and we look forward to seeing more of your work.
Summary
A passive background daemon that analyzes HOW the user communicates, not just WHAT they say. Tracks filler words, hedging language, vocabulary diversity, question-to-statement ratio, and utterance length. Delivers on-demand spoken coaching reports with historical trends, personal milestones, and actionable improvement tips.
Features
Testing
All analysis happens passively in the background while you talk naturally.