AGENT SKILL

granola-performance-tuning

Optimize Granola transcription quality and note performance. Use when improving transcription accuracy, reducing processing time, or enhancing note quality. Trigger with phrases like "granola performance", "granola accuracy", "granola quality", "improve granola", "granola optimization".

License 许可
MIT
最近更新
2026-11-15
Stars
★ 2.0k
Forks
233

01 前言 Front Matter

name
granola-performance-tuning
author
Jeremy Longshore <jeremy@intentsolutions.io>
license
MIT
version
1.0.0
description
Optimize Granola transcription quality and note performance. Use when improving transcription accuracy, reducing processing time, or enhancing note quality. Trigger with phrases like "granola performance", "granola accuracy", "granola quality", "improve granola", "granola optimization".
allowed-tools
Read, Write, Edit

03 SKILL.md 正文

Granola Performance Tuning

Overview

Optimize Granola for best transcription accuracy and note quality.

Transcription Quality Factors

Audio Quality Hierarchy

Transcription Accuracy
        ↑
[Professional Microphone] 98%
        ↑
[Quality Headset Mic] 95%
        ↑
[Laptop Built-in Mic] 85%
        ↑
[Phone Speaker] 70%

Environmental Factors

Factor Impact Optimization
Background noise High Use quiet room, noise cancellation
Echo/reverb High Soft furnishings, smaller room
Distance from mic Medium Within 12 inches of microphone
Multiple speakers Medium Use identification phrases
Accent variation Low Improves over time with usage

Audio Setup Optimization

Recommended Equipment

## Microphone Recommendations

Budget (~$50):
- Blue Snowball iCE
- Fifine K669

Mid-Range (~$100):
- Blue Yeti
- Rode NT-USB Mini
- Audio-Technica AT2020USB+

Professional (~$200+):
- Shure MV7
- Elgato Wave:3
- Rode PodMic + interface

Microphone Settings (macOS)

# Check current input device
system_profiler SPAudioDataType | grep -A5 "Default Input"

# Adjust input volume (System Preferences)
# Aim for: Input level peaks at 75% during normal speech

Room Optimization

## Environment Checklist
- [ ] Close windows to reduce outside noise
- [ ] Turn off fans, AC if possible
- [ ] Use soft surfaces (carpet, curtains)
- [ ] Position away from keyboard clicks
- [ ] Mute when not speaking

Note Quality Optimization

Meeting Preparation

## Pre-Meeting Checklist
- [ ] Share agenda in advance
- [ ] Send attendee list to calendar
- [ ] Prepare context notes in template
- [ ] Test audio before meeting

During Meeting

## Best Practices
1. State names when addressing people
   "Sarah, what do you think about..."

2. Summarize decisions verbally
   "So we're agreed: deadline is Friday."

3. Spell out technical terms
   "The API endpoint, A-P-I..."

4. Avoid crosstalk
   One person speaking at a time

5. Use clear action item language
   "Action item: Mike will review the PR by Thursday."

Post-Meeting Enhancement

## Note Review Checklist (5 min)
- [ ] Correct obvious transcription errors
- [ ] Add context AI might have missed
- [ ] Verify action items are complete
- [ ] Add links to referenced documents
- [ ] Tag key decisions

Template Optimization

Effective Template Structure

# Meeting Template: Sprint Planning

## Agenda (Pre-filled)
-

## Context
[Add links to relevant docs]

## Discussion Notes
[AI-enhanced during meeting]

## Decisions
- [ ] Decision 1: [Clear statement]

## Action Items
Format: - [ ] What (@who, by when)

## Follow-up
Next meeting: [date]

Template Best Practices

Practice Reason Impact
Use headers Better AI parsing +20% accuracy
Pre-fill context Reduces ambiguity +15% relevance
Standard formats Consistent output +10% usability
Action item format Auto-extraction +25% detection

Processing Speed Optimization

Factors Affecting Speed

Factor Impact Optimization
Meeting length Linear Expect 1 min processing per 10 min meeting
Internet speed High Ensure stable connection during upload
Peak times Medium Processing queue varies
Audio quality Low Cleaner audio = faster processing

Speed Expectations

Meeting Duration → Processing Time
15 minutes → 1-2 minutes
30 minutes → 2-3 minutes
60 minutes → 3-5 minutes
120 minutes → 5-8 minutes

Integration Performance

Zapier Optimization

## Reduce Zapier Latency

1. Use Instant triggers (not polling)
2. Minimize steps in Zap
3. Avoid unnecessary filters
4. Use multi-step Zaps efficiently
5. Monitor task usage

Batch Processing

# Instead of real-time, batch for efficiency
Schedule: Every 30 minutes
Process:
  - Collect all new notes
  - Batch update Notion
  - Single Slack summary
  - Aggregate CRM updates

Accuracy Improvement

Training the AI

## Improve Over Time

1. Correct errors when you see them
   - AI learns from corrections

2. Use consistent terminology
   - Builds vocabulary

3. Identify speakers
   - Improves attribution

4. Regular editing
   - Provides feedback loop

Custom Vocabulary

## Teach Domain Terms

Add to meeting intros:
"We'll discuss the OAuth2 implementation,
that's O-Auth-Two, and the GraphQL API,
spelled G-R-A-P-H-Q-L..."

Common terms to spell out:
- Acronyms (API, SDK, CI/CD)
- Product names
- People names with unusual spellings

Performance Metrics

What to Track

Metric Target How to Measure
Transcription accuracy >95% Sample review
Action item detection >90% Compare to meeting
Processing time <5 min Timestamp comparison
Note usefulness 4+/5 Team survey

Weekly Review

## Performance Check

Monday:
- [ ] Review last week's meeting notes
- [ ] Note common transcription errors
- [ ] Identify improvement opportunities
- [ ] Adjust templates if needed

Resources

Next Steps

Proceed to granola-cost-tuning for cost optimization strategies.

04 同仓库更多 Skills

Dicklesworthstone
Dicklesworthstone/pi_agent_rust

theme-factory

Toolkit for styling artifacts with a theme. These artifacts can be slides, docs, reportings, HTML landing pages, etc. There are 10 pre-set themes with colors/fonts that you can apply to any artifact that has been creating, or can generate a new theme on-the-fly.

视觉设计
★ 2.0k
Dicklesworthstone
Dicklesworthstone/pi_agent_rust

web-artifacts-builder

Suite of tools for creating elaborate, multi-component claude.ai HTML artifacts using modern frontend web technologies (React, Tailwind CSS, shadcn/ui). Use for complex artifacts requiring state management, routing, or shadcn/ui components - not for simple single-file HTML/JSX artifacts.

前端开发
★ 2.0k
Dicklesworthstone
Dicklesworthstone/pi_agent_rust

webapp-testing

Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.

测试与质量前端开发
★ 2.0k
Dicklesworthstone
Dicklesworthstone/pi_agent_rust

neuropixels-analysis

Neuropixels neural recording analysis. Load SpikeGLX/OpenEphys data, preprocess, motion correction, Kilosort4 spike sorting, quality metrics, Allen/IBL curation, AI-assisted visual analysis, for Neuropixels 1.0/2.0 extracellular electrophysiology. Use when working with neural recordings, spike sorting, extracellular electrophysiology, or when the user mentions Neuropixels, SpikeGLX, Open Ephys, Kilosort, quality metrics, or unit curation.

★ 2.0k
Dicklesworthstone
Dicklesworthstone/pi_agent_rust

skill-creator

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

Skill与智能体开发
★ 2.0k
Dicklesworthstone
Dicklesworthstone/pi_agent_rust

notebooklm

Use this skill to query your Google NotebookLM notebooks directly from Claude Code for source-grounded, citation-backed answers from Gemini. Browser automation, library management, persistent auth. Drastically reduced hallucinations through document-only responses.

Skill与智能体开发
★ 2.0k