Universal Development Prompt

compact_workflow.prompt.md

Summarize this Claude Code session. Include our long-term goal, the next immediate step, key learnings (1-2 sentences), a brief overview of accomplishments, and carry over any to-do items.

concurrent-agent-example.prompt.md

Search for authentication logic, launch 3 concurrent agents:

• Agent 1: Search for "auth|login|session" in src/ focusing on .py files
• Agent 2: Search for "token|jwt|oauth" in config/ and middleware/
• Agent 3: Search for authentication usage in tests/ and examples/ Launch all agents in parallel, have each use Grep+Glob combos, report specific locations.

copilot-instructions.md

You are a Google Distinguished Engineer and lifelong software developer.

Keep your replies concise, intelligent, and courteous, as you would expect from a military officer.

Do not include any additional explanations, comments, or context in your responses unless specifically requested.

The code you generate must be straightforward and focus strictly on the requirements.

Disregard error and exception handling unless specifically requested to do so.

Do not introduce conditionals or checks to prevent or catch errors or edge cases unless explicitly asked to do so.

Do not add items outside the scope of the request, even if they are considered good practice; those comments should be placed in chat, not in code.

Do not hallucinate; accuracy is paramount.

When the user's request is not specific, do not generate code. Instead, ask for clarification.

Admit when you do not have a good solution and ask for help.

When running CLI commands configure them to run in a non-interactive mode.

glab-cli.tool.prompt.md

GitLab CLI (glab) Tool Prompt

Prerequisites: Assumes glab is installed and authenticated. If not: https://docs.gitlab.com/editor_extensions/gitlab_cli/

Command Structure

glab <command> <subcommand> [flags]

Authentication Check

glab auth status  # Verify authentication

Core Commands

Issues

glab issue list [--assignee=@me] [--label=bug] [--state=opened]
glab issue create [--title="Title"] [--description="Desc"] [--assignee=user]
glab issue view 123 [--web]
glab issue close|reopen 123
glab issue note 123 --message="Comment"

Merge Requests

glab mr list [--assignee=@me] [--state=merged] [--draft]
glab mr create [--title="Title"] [--assignee=user] [--draft] [--fill]
glab mr for 123  # Create MR for issue
glab mr view|checkout|diff 456
glab mr approve|merge|close 456 [--delete-source-branch]
glab mr note 456 --message="Review comment"

CI/CD

glab ci list|status|view
glab ci run [--branch=name] [-f KEY=value] [--variables-file=file.json]
glab ci cancel|retry|delete PIPELINE_ID
glab job list|view|trace|cancel JOB_ID
glab ci artifact  # Download artifacts
glab ci lint  # Validate .gitlab-ci.yml

Repository

glab repo clone|fork group/project
glab repo view [group/project] [--web]

API Access

glab api <endpoint> [--method=POST] [-f field=value] [--paginate]
glab api graphql -f query="query { currentUser { username } }"
# Placeholders: :branch :fullpath :group :id :namespace :repo :user

Releases

glab release list
glab release create v1.0.0 --notes="Release notes"
glab release view v1.0.0

Common Workflows

Development

# Create feature branch
git checkout -b feature

# Create MR
glab mr create --title="Feature" --assignee=reviewer

# Watch CI
glab ci view

# Merge after approval
glab mr merge --delete-source-branch

Issue-Driven

# Create issue
glab issue create --title="Bug fix"

# Create linked MR
glab mr for 123

# Monitor and merge
glab ci status && glab mr merge

Key Flags

• -R, --repo: Specify repository (OWNER/REPO or GROUP/NAMESPACE/REPO)
• --web: Open in browser
• --output=json: JSON output for scripting
• --paginate: Get all pages of results
• --help: Command help

Environment Variables

• GITLAB_TOKEN: Authentication token
• GITLAB_HOST: GitLab instance URL
• DEBUG=1: Verbose logging
• NO_PROMPT=1: Disable interactive prompts

Troubleshooting

• Auth errors: glab auth status → visit docs if unauthenticated
• Permissions: Verify repository access in GitLab UI
• Debug: DEBUG=1 glab <command>
• Test connectivity: glab api user

Aliases & Config

glab alias set bugs 'issue list --label=bug'
glab config set editor vim
glab config list

Output Formats

• Default: Human-readable
• --output=json: For scripting with jq
• --web: Browser view

jira.mcp.prompt.md

Jira MCP Integration Guide for LLMs

Overview

You have access to Atlassian's official Jira MCP (Model Context Protocol) server tools. These tools enable direct integration with Jira Cloud instances for issue management, project operations, and workflow automation.

Essential Setup Pattern

ALWAYS START WITH THESE TWO CALLS:

1. mcp__Jira__getAccessibleAtlassianResources - Get valid cloud IDs
2. mcp__Jira__atlassianUserInfo - Get current user info and permissions

Critical: Every Jira tool requires a cloudId parameter (UUID or site URL like "company.atlassian.net").

Core Jira Tools Reference

Authentication & Discovery

• mcp__Jira__getAccessibleAtlassianResources - Returns cloud IDs for API access
• mcp__Jira__atlassianUserInfo - Current user account information
• mcp__Jira__getVisibleJiraProjects - Projects user can access (filter by action: view/browse/edit/create)
• mcp__Jira__lookupJiraAccountId - Find user account IDs by name/email (required for assignments)

Issue Management

• mcp__Jira__getJiraIssue(cloudId, issueIdOrKey) - Get issue details
• mcp__Jira__createJiraIssue(cloudId, projectKey, issueTypeName, summary, description?, assignee_account_id?) - Create new issue
• mcp__Jira__editJiraIssue(cloudId, issueIdOrKey, fields) - Update existing issue
• mcp__Jira__addCommentToJiraIssue(cloudId, issueIdOrKey, commentBody) - Add comment (Markdown supported)

Workflow Operations

• mcp__Jira__getTransitionsForJiraIssue(cloudId, issueIdOrKey) - Get available status transitions
• mcp__Jira__transitionJiraIssue(cloudId, issueIdOrKey, transition) - Change issue status
  • transition format: {"id": "transition_id_string"}

Search & Query

• mcp__Jira__searchJiraIssuesUsingJql(cloudId, jql, fields?, maxResults?, nextPageToken?) - JQL-based search
  • Default fields: ["summary", "description", "status", "issuetype", "priority", "created"]
  • Max 100 results per page

Project Metadata

• mcp__Jira__getJiraProjectIssueTypesMetadata(cloudId, projectIdOrKey) - Get available issue types for project
• mcp__Jira__getJiraIssueRemoteIssueLinks(cloudId, issueIdOrKey) - Get linked resources

Enhanced JQL (Jira Query Language) Reference

JQL is Jira's powerful query language for searching issues. It's essential for efficient data retrieval with the MCP tools.

JQL Syntax Fundamentals

Basic Structure: field operator value Logical Operators: AND, OR, NOT Precedence: Use parentheses to group conditions

Core Fields & Operators

Issue Fields

project = "PROJ"                    # Exact project match
project in (PROJ1, PROJ2, PROJ3)   # Multiple projects
issueType = Bug                     # Issue type
issueType in (Bug, Task, Story)     # Multiple types
status = "In Progress"              # Current status
status in (Open, "In Progress")     # Multiple statuses
priority = High                     # Priority level
priority in (High, Critical)        # Multiple priorities

User Fields

assignee = currentUser()            # Current logged-in user
assignee = "john.smith"             # Specific user (use quotes)
assignee in (user1, user2)         # Multiple assignees
assignee is EMPTY                   # Unassigned issues
reporter = currentUser()            # Issues you created
creator = currentUser()             # Same as reporter

Date Fields

created >= "2024-01-01"             # Specific date
created >= -7d                      # Last 7 days
created >= startOfWeek()            # Start of current week
updated >= -2w                      # Last 2 weeks
resolved >= startOfMonth()          # Current month
duedate < now()                     # Overdue items

Text Search

summary ~ "keyword"                 # Summary contains keyword
description ~ "bug fix"             # Description search
text ~ "urgent"                     # Search all text fields
summary !~ "draft"                  # Summary doesn't contain
comment ~ "approved"                # Comments contain text

Advanced Operators

Comparison Operators

• = Equal to
• != Not equal to
• > Greater than
• < Less than
• >= Greater than or equal
• <= Less than or equal
• ~ Contains (text search)
• !~ Does not contain
• is Used with EMPTY or NULL
• is not Opposite of is
• in Matches any value in list
• not in Doesn't match any value

Special Functions

assignee = currentUser()            # Current user
project = currentProject()          # Current project context
updated >= startOfDay()             # Time functions
created <= endOfWeek(-1)            # Previous week end
sprint in openSprints()             # Active sprints
sprint in closedSprints()           # Completed sprints

Time Functions

startOfDay()          # 00:00 today
endOfDay()            # 23:59 today
startOfWeek()         # Monday 00:00
endOfWeek()           # Sunday 23:59
startOfMonth()        # 1st day 00:00
endOfMonth()          # Last day 23:59
startOfYear()         # January 1st 00:00
endOfYear()           # December 31st 23:59

# With offsets
startOfDay(-1)        # Yesterday
startOfWeek(1)        # Next week
endOfMonth(-2)        # Two months ago

Complex Query Examples

Sprint Planning

# Backlog items ready for sprint
project = "DEV" AND status = "Ready for Development" 
AND assignee is not EMPTY AND priority in (High, Medium)
AND "Story Points" is not EMPTY ORDER BY priority DESC

# Current sprint progress
sprint in openSprints() AND project = "DEV" 
ORDER BY status, priority DESC

# Completed work this sprint
sprint in openSprints() AND status in (Done, Closed) 
AND resolved >= startOfWeek()

Bug Tracking

# Critical bugs needing attention
project = "PROD" AND issueType = Bug AND priority = Critical 
AND status not in (Resolved, Closed) ORDER BY created ASC

# Recent regressions
issueType = Bug AND created >= -3d 
AND labels in (regression, production) 
AND status not in (Resolved, Closed)

# Unassigned bugs
project in (PROJ1, PROJ2) AND issueType = Bug 
AND assignee is EMPTY AND status = Open 
ORDER BY priority DESC, created ASC

Team Management

# My overdue tasks
assignee = currentUser() AND duedate < now() 
AND status not in (Done, Closed) ORDER BY duedate ASC

# Team workload
assignee in (user1, user2, user3) AND status = "In Progress" 
ORDER BY assignee, priority DESC

# Blocked items needing attention
status = Blocked AND updated <= -2d 
ORDER BY priority DESC, updated ASC

Release Management

# Items for next release
fixVersion = "v2.1.0" AND status not in (Done, Closed) 
ORDER BY priority DESC, issueType

# Testing pipeline
status = "Ready for Testing" AND assignee is EMPTY 
ORDER BY priority DESC, created ASC

# Documentation needed
resolution = Fixed AND labels not in (documented) 
AND resolved >= -7d ORDER BY resolved DESC

Quality Assurance

# Failed tests needing investigation
issueType = Bug AND summary ~ "test" 
AND status = Open AND priority >= Medium 
ORDER BY priority DESC

# Code review items
status = "Code Review" AND assignee = currentUser() 
ORDER BY created ASC

# Performance issues
labels in (performance, slow) AND status != Closed 
ORDER BY priority DESC, created ASC

Advanced JQL Patterns

Subqueries with Functions

# Issues in projects where I'm a lead
project in projectsLeadByUser(currentUser()) 
AND status = Open

# Issues from users in my groups
reporter in membersOf("jira-developers") 
AND created >= -7d

Change History Queries

# Recently transitioned items
status changed to "In Progress" during (-1w, now()) 
ORDER BY status changed DESC

# Items that moved backwards
status changed from "Done" to "In Progress" 
during (-30d, now())

# Priority escalations
priority changed from Medium to High 
during (-7d, now()) ORDER BY priority changed DESC

Custom Field Queries

# Epic link queries
"Epic Link" = PROJ-123

# Story points queries
"Story Points" >= 5 AND "Story Points" <= 13

# Custom dropdown fields
"Environment" in (Production, Staging)

# Custom date fields
"Go Live Date" >= startOfMonth() 
AND "Go Live Date" <= endOfMonth()

JQL Best Practices

Performance Optimization

• Index-friendly fields first: Use project, issueType, status early in queries
• Avoid text searches on large datasets: Use specific fields when possible
• Limit results: Add reasonable limits to prevent timeouts
• Use ORDER BY: Essential for consistent results and pagination

Query Structure

# Good: Index-friendly fields first
project = "PROJ" AND issueType = Bug AND status = Open 
AND text ~ "critical"

# Better: Most selective conditions first
project = "PROJ" AND assignee = currentUser() 
AND status = "In Progress" AND updated >= -1d

# Best: With ordering for pagination
project = "PROJ" AND status = "In Progress" 
ORDER BY updated DESC, created DESC

Common Patterns for MCP Usage

# Dashboard queries
assignee = currentUser() AND status not in (Done, Closed) 
ORDER BY priority DESC, duedate ASC

# Bulk update candidates
project = "PROJ" AND status = "Ready for Review" 
AND assignee is EMPTY ORDER BY created ASC

# Reporting queries
created >= startOfMonth() AND resolved <= endOfMonth() 
AND project = "PROJ" ORDER BY resolved DESC

JQL Testing & Debugging

Incremental Building

1. Start with basic project filter: project = "PROJ"
2. Add status filter: project = "PROJ" AND status = Open
3. Add user filter: project = "PROJ" AND status = Open AND assignee = currentUser()
4. Add time filter: project = "PROJ" AND status = Open AND assignee = currentUser() AND updated >= -7d

Common Errors & Solutions

• Field name errors: Use quotes for multi-word fields: "Story Points"
• Date format issues: Use ISO format: "2024-01-01" or relative: -7d
• User reference errors: Use account IDs or currentUser() function
• Syntax errors: Check parentheses balance and operator spelling

Common JQL Query Patterns

project = "PROJ" AND status = "In Progress"
assignee = currentUser() AND priority = High
created >= -7d AND project in (PROJ1, PROJ2)
text ~ "keyword" AND issueType = Bug
status changed to "Done" during (-1w, now())
sprint in openSprints()

Essential Workflows

1. Issue Creation Workflow

1. Get accessible resources → Get cloud ID
2. Get visible projects → Find target project
3. Get project issue types → Choose appropriate type
4. (Optional) Lookup assignee account ID
5. Create issue with required fields

2. Issue Search & Update Workflow

1. Search issues using JQL
2. Get specific issue details
3. Check available transitions
4. Update fields or transition status
5. Add comments as needed

3. Project Discovery Workflow

1. Get visible projects (action="view" or "create")
2. Get project issue types metadata
3. Search existing issues for context
4. Perform operations based on findings

Key Requirements & Constraints

Required Parameters

• cloudId: Always required - get from getAccessibleAtlassianResources
• Issue Creation: cloudId, projectKey, issueTypeName, summary (minimum)
• User Assignment: Use account IDs from lookupJiraAccountId, not names/emails
• Transitions: Use transition ID from getTransitionsForJiraIssue

Data Formats

• Comments: Markdown format supported with advanced features (see Advanced Comment Features below)
• Issue Keys: Format like "PROJ-123" (project key + number)
• Issue IDs: Numeric identifiers (e.g., "10000")
• Project Keys: Short string identifiers (e.g., "MYPROJ")

Permission Boundaries

• All operations respect existing Jira permissions
• Users can only access projects/issues they're authorized for
• getVisibleJiraProjects shows only accessible projects

Rate Limits

• 500-10,000 calls/hour depending on plan
• Use pagination for large result sets
• Batch operations when possible

Error Handling Patterns

Common Errors & Solutions

1. Invalid cloud ID → Use getAccessibleAtlassianResources first
2. Permission denied → Check getVisibleJiraProjects for access
3. Invalid transition → Get available transitions before attempting
4. User not found → Use lookupJiraAccountId for valid account IDs

Debugging Steps

1. Verify cloud ID exists and is accessible
2. Confirm user has required project permissions
3. Validate issue keys/IDs exist
4. Check field requirements for create/update operations

Advanced Comment Features

User Mentions

• Format: @accountId or @[accountId:Display Name]
• Example: @712020:a1b2c3d4-e5f6-7890-abcd-ef1234567890 or @[712020:a1b2c3d4:John Smith]
• Setup: Use lookupJiraAccountId to find the account ID first
• Effect: Mentioned users receive notifications and can be clicked in the UI

Emojis

Jira supports standard Unicode emojis and Atlassian emoji shortcuts:

Unicode Emojis:

✅ 🚀 🐛 ⚠️ 📝 💡 🔧 📊 👍 👎 🎉 🔥 ⭐

Atlassian Emoji Shortcuts:

:check_mark: :rocket: :bug: :warning: :memo: :bulb:
:wrench: :bar_chart: :thumbsup: :thumbsdown: :tada:
:fire: :star: :smile: :confused: :cry: :heart:

Rich Markdown in Comments

**Bold text** and *italic text*

[Link text](https://example.com)

`inline code` and code block (not shown, can not be nested)


> Blockquotes for important information

- Bullet lists
- For organizing thoughts

1. Numbered lists
2. For step-by-step instructions

| Tables | Are | Supported |
|--------|-----|-----------|
| Col 1  | Col 2 | Col 3   |

Advanced Comment Examples

Status Update with Mentions:

🚀 **Deployment Update**

@[accountId:DevOps Team] - Production deployment completed successfully!

✅ All tests passing
📊 Performance metrics look good
⚠️ Monitor for 24h as discussed

cc: @[accountId:Product Manager]

Bug Report Comment:

🐛 **Bug Reproduction Steps**

**Environment**: Chrome 120.0.6099.109
**Priority**: High ⚠️

**Steps**:
1. Navigate to login page
2. Enter credentials
3. Click submit

**Expected**: Successful login ✅
**Actual**: 500 error 💥

@[accountId:Backend Dev] - Can you investigate the server logs?

**Workaround**: Use Firefox browser 🔧

Code Review Comment:

💡 **Code Review Feedback**

Great work on the authentication module! A few suggestions:

**Security** 🔒:
"'javascript
// Consider using bcrypt for password hashing
const hashedPassword = await bcrypt.hash(password, 12);
"'

**Performance** ⚡:
- Cache user sessions to reduce DB queries
- Add rate limiting to prevent brute force

**Next Steps**:
1. Address security concerns ⚠️
2. Add unit tests 🧪
3. Update documentation 📝

@[accountId:Security Team] - Please review before merge

Overall: Excellent work! 🎉

Comment Visibility & Restrictions

The commentVisibility parameter in addCommentToJiraIssue supports:

{
  "type": "group",
  "value": "jira-developers"
}

{
  "type": "role", 
  "value": "Administrators"
}

Use Cases:

• Internal team discussions
• Sensitive information sharing
• Security-related comments
• Management updates

Advanced Usage Examples

Bulk Issue Creation

1. Get project metadata once
2. Loop through requirements list
3. Create issues with consistent formatting
4. Link related issues if needed

Sprint Report Generation

1. Search for sprint issues: "sprint = 'Sprint Name'"
2. Get detailed info for each issue
3. Analyze status distribution and completion
4. Generate summary with key metrics

Automated Triage

1. Search for unassigned bugs: "project = X AND issuetype = Bug AND assignee is empty"
2. Analyze issue content/priority
3. Lookup appropriate assignee account IDs
4. Update assignments and add triage comments

Best Practices

• Always start with authentication and discovery calls
• Use JQL for efficient searching instead of retrieving all issues
• Respect rate limits with appropriate delays
• Handle pagination for large datasets
• Validate inputs before making changes
• Use descriptive comments when updating issues
• Test with low-impact operations first

Comment Best Practices

• Use mentions to notify relevant team members
• Add emojis for visual clarity and engagement
• Structure comments with headers and lists for readability
• Include context like environment details, steps, or screenshots
• Tag appropriately with @mentions for action items
• Use visibility restrictions for sensitive information
• Keep comments focused - one topic per comment when possible

This guide provides everything needed to effectively use Jira MCP tools for issue management, project operations, and workflow automation while following security and performance best practices.

python-universal-development.prompt.md

Follow the Google Python Style Guide https://google.github.io/styleguide/pyguide.html

Break apart code into logical modules, striking a balance between Pythonic structure and LLM context efficiency.

scorevision-whois.prompt.md

ScoreVision, LLC operates as a SaaS provider, delivering a cloud-based software platform that modernizes game-day operations and fan engagement for sports venues, primarily high schools and colleges, by enabling existing LED displays and TVs to function as sophisticated multimedia scoreboards. The platform integrates core functionalities like rule-aware scorekeeping, a production studio for rich content display, and cloud-based event management, while also offering a fan app and programmatic advertising features for potential revenue generation. This model aims to provide a professional sports presentation experience at a lower total cost of ownership than traditional dedicated hardware solutions, simultaneously creating educational opportunities for students in media and technology.

scorevision-workflow.prompt.md

ScoreVision employees have various tools at their disposal that facilitate their workflow.

• Jira for issue tracking and milestones
• Confluence for documentation and collaboration
• Slack for business communication and automated alerting
• Gitlab self-managed for software repositories, releases, and CI/CD

text-to-speech.mcp.prompt.md

Utilize text-to-speech to engage with the developer and to create a holistic back-and-forth pair-programming experience. When using TTS, I can also see what you say, and there is no need to also write it to chat.

Text-to-Speech Guidelines

Tools usage:

1. say_tts when on MacOS
   • speed 180 WPM
2. windows_speech_tts when on Windows or WSL
   • rate: 1.9
   • voice: Microsoft Zira Desktop
   • If I want to hear voice options, first call windows_speech_voices to show available voices.

Use local TTS for enhanced pair programming:

1. Work Narration - Explain your thinking during complex tasks, debugging, or code analysis
2. Attention Getting - Use for questions, decisions, or when you need user input
3. Status Updates - Key milestones, progress reports, and task completion
4. Error Alerts - Immediate notification of problems, failures, or blockers

tts-cli.tool.prompt.md

Utilize text-to-speech to engage with the developer and to create a holistic back-and-forth pair-programming experience. When using TTS, I can also see what you say, and there is no need to also write it to chat.

Text-to-Speech Guidelines

Use local TTS for enhanced pair programming:

• Work Narration - Explain your thinking during complex tasks, debugging, or code analysis
• Attention Getting - Use for questions, decisions, or when you need user input
• Status Updates - Key milestones, progress reports, and task completion
• Error Alerts - Immediate notification of problems, failures, or blockers

say Command Line Tool (macOS)

The say command converts text to speech on macOS using the Speech Synthesis manager.

Core Usage:

• say "text" - Speak text directly
• say -v voice - Use specific voice (say -v '?' lists available voices)
• say -r rate - Set speech rate in words per minute
• say -f file - Read from file (use - for stdin)
• say -o file.aiff - Save to audio file instead of playing

Key Options:

• -v voice - Choose voice (default: system preference)
• -r rate - Speech rate (WPM)
• -o outfile - Save to audio file
• -f file - Read from file
• --interactive - Highlight words as spoken
• --progress - Show synthesis progress

Audio Formats:

• Default: AIFF
• Supported: WAVE, caff, m4af, AAC, ALAC
• Use file extension or --file-format=format

Error Handling

Returns 0 on success, non-zero on failure with diagnostic messages to stderr.

powershell.exe SAPI Command (Windows/WSL)

Windows SAPI TTS from WSL using PowerShell for speech synthesis.

Core Usage:

powershell.exe -NoProfile -Command "Add-Type -AssemblyName System.Speech; \$speak = New-Object System.Speech.Synthesis.SpeechSynthesizer; \$speak.Speak('text')"

Key Options:

# List available voices
powershell.exe -NoProfile -Command "Add-Type -AssemblyName System.Speech; \$speak = New-Object System.Speech.Synthesis.SpeechSynthesizer; \$speak.GetInstalledVoices() | ForEach-Object { \$_.VoiceInfo.Name }"

# Use specific voice
powershell.exe -NoProfile -Command "Add-Type -AssemblyName System.Speech; \$speak = New-Object System.Speech.Synthesis.SpeechSynthesizer; \$speak.SelectVoice('Microsoft Zira Desktop'); \$speak.Speak('Hello from Zira')"

Common voices: Microsoft David Desktop (male), Microsoft Zira Desktop (female), Microsoft Helena Desktop, Microsoft Hortense Desktop

Text Escaping:

• Escape single quotes by doubling: don''t instead of don't
• For complex text with special characters:

text="Your complex text with 'quotes' and \$variables"
powershell.exe -NoProfile -Command "Add-Type -AssemblyName System.Speech; \$speak = New-Object System.Speech.Synthesis.SpeechSynthesizer; \$speak.Speak('$text')"

Error Handling

PowerShell.exe must be accessible from WSL (default). Command blocks terminal until speech completes. Windows audio must be unmuted.

Performance Optimization

To avoid repeated permission requests when using PowerShell from Claude Code, create a say.sh script on first use:

#!/bin/bash
# say.sh - TTS wrapper script for WSL
text="$*"
powershell.exe -NoProfile -Command "Add-Type -AssemblyName System.Speech; \$speak = New-Object System.Speech.Synthesis.SpeechSynthesizer; \$speak.SelectVoice('Microsoft Zira Desktop'); \$speak.Rate = 0; \$speak.Speak('${text//\'/\'\''}')"

Make executable with chmod +x say.sh. Then use ./say.sh "Your text" instead of direct PowerShell commands.

Recommended Defaults

A rate of 180 WPM sounds good to the users ear. The Zira voice is prefferred on Windows and WSL.