Ai Keeper/Guides/Every Work Mode
← All workflows
Collection 02 · Outcome presets

Every Work Mode

Choose the exact outcome you want, review what that preset changes, satisfy only its real requirements, and prove the starter behavior before using important data.

67 workflowsBeginner-first

Find one workflow

Filter this collection by name, outcome, provider, engine, channel, mode, or command.

67 shown
Workflow 01

Use the Secure Coder Work Mode

Safer code changes, threat awareness, regression discipline, and minimal blast radius. This mode belongs to the Developer Copilot family.

Beginner5–15 minResult · Secure Coder is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Current Work Modes hub with family grouping and Start Here actions.
Every Work Mode is applied from this hub. The card shows its family before you commit to it.
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Developer Copilot family, and select Secure Coder.

  2. Read the exact outcome before applying it: Safer code changes, threat awareness, regression discipline, and minimal blast radius.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Explain the safest first change and how you would verify it. Approve no mutation unless the preview matches your intent.

You are done when
Secure Coder is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 02

Use the Full-Stack Builder Work Mode

Ships features across UI, backend, tooling, and runtime wiring. This mode belongs to the Developer Copilot family.

Beginner5–15 minResult · Full-Stack Builder is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Developer Copilot family, and select Full-Stack Builder.

  2. Read the exact outcome before applying it: Ships features across UI, backend, tooling, and runtime wiring.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Explain the safest first change and how you would verify it. Approve no mutation unless the preview matches your intent.

You are done when
Full-Stack Builder is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 03

Use the Code Explainer Work Mode

Explains architecture, files, bugs, and tradeoffs without over-editing. This mode belongs to the Developer Copilot family.

Beginner5–15 minResult · Code Explainer is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Developer Copilot family, and select Code Explainer.

  2. Read the exact outcome before applying it: Explains architecture, files, bugs, and tradeoffs without over-editing.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Explain the safest first change and how you would verify it. Approve no mutation unless the preview matches your intent.

You are done when
Code Explainer is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 04

Use the Terminal Assistant Work Mode

Helps with shell commands, local debugging, and CLI ergonomics. This mode belongs to the Developer Copilot family.

Beginner5–15 minResult · Terminal Assistant is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Developer Copilot family, and select Terminal Assistant.

  2. Read the exact outcome before applying it: Helps with shell commands, local debugging, and CLI ergonomics.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Explain the safest first change and how you would verify it. Approve no mutation unless the preview matches your intent.

You are done when
Terminal Assistant is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 05

Use the Git Commit Writer Work Mode

Turns diffs and context into clear commits, PR notes, and changelogs. This mode belongs to the Developer Copilot family.

Beginner5–15 minResult · Git Commit Writer is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Developer Copilot family, and select Git Commit Writer.

  2. Read the exact outcome before applying it: Turns diffs and context into clear commits, PR notes, and changelogs.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Explain the safest first change and how you would verify it. Approve no mutation unless the preview matches your intent.

You are done when
Git Commit Writer is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 06

Use the Refactoring Agent Work Mode

Untangles large code paths while preserving behavior and readability. This mode belongs to the Developer Copilot family.

Beginner5–15 minResult · Refactoring Agent is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Developer Copilot family, and select Refactoring Agent.

  2. Read the exact outcome before applying it: Untangles large code paths while preserving behavior and readability.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Explain the safest first change and how you would verify it. Approve no mutation unless the preview matches your intent.

You are done when
Refactoring Agent is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 07

Use the Test Generator Work Mode

Adds regression tests, fixtures, and coverage-oriented checks. This mode belongs to the Developer Copilot family.

Beginner5–15 minResult · Test Generator is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Developer Copilot family, and select Test Generator.

  2. Read the exact outcome before applying it: Adds regression tests, fixtures, and coverage-oriented checks.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Explain the safest first change and how you would verify it. Approve no mutation unless the preview matches your intent.

You are done when
Test Generator is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 08

Use the Migration Assistant Work Mode

Plans and executes framework, API, schema, and tooling migrations. This mode belongs to the Developer Copilot family.

Beginner5–15 minResult · Migration Assistant is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Developer Copilot family, and select Migration Assistant.

  2. Read the exact outcome before applying it: Plans and executes framework, API, schema, and tooling migrations.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Explain the safest first change and how you would verify it. Approve no mutation unless the preview matches your intent.

You are done when
Migration Assistant is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 09

Use the Research Assistant Work Mode

Fast, structured research with citations, synthesis, and takeaways. This mode belongs to the Research & Knowledge family.

Beginner5–15 minResult · Research Assistant is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Research & Knowledge family, and select Research Assistant.

  2. Read the exact outcome before applying it: Fast, structured research with citations, synthesis, and takeaways.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Create a source-backed outline and mark uncertainty. Approve no mutation unless the preview matches your intent.

You are done when
Research Assistant is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 10

Use the Deep Researcher Work Mode

Long-form investigation across multiple sources, comparisons, and source trails. This mode belongs to the Research & Knowledge family.

Beginner5–15 minResult · Deep Researcher is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Research & Knowledge family, and select Deep Researcher.

  2. Read the exact outcome before applying it: Long-form investigation across multiple sources, comparisons, and source trails.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Create a source-backed outline and mark uncertainty. Approve no mutation unless the preview matches your intent.

You are done when
Deep Researcher is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 11

Use the Knowledge Base Chat Work Mode

Turns local memory and indexed documents into a reusable answer surface. This mode belongs to the Research & Knowledge family.

Beginner5–15 minResult · Knowledge Base Chat is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Research & Knowledge family, and select Knowledge Base Chat.

  2. Read the exact outcome before applying it: Turns local memory and indexed documents into a reusable answer surface.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Create a source-backed outline and mark uncertainty. Approve no mutation unless the preview matches your intent.

You are done when
Knowledge Base Chat is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 12

Use the Paper Analyzer Work Mode

Understands research papers, methods, claims, and limitations. This mode belongs to the Research & Knowledge family.

Beginner5–15 minResult · Paper Analyzer is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Research & Knowledge family, and select Paper Analyzer.

  2. Read the exact outcome before applying it: Understands research papers, methods, claims, and limitations.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Create a source-backed outline and mark uncertainty. Approve no mutation unless the preview matches your intent.

You are done when
Paper Analyzer is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 14

Use the Meeting Transcript Processor Work Mode

Turns transcripts into summaries, actions, owners, and knowledge notes. This mode belongs to the Research & Knowledge family.

Beginner5–15 minResult · Meeting Transcript Processor is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Research & Knowledge family, and select Meeting Transcript Processor.

  2. Read the exact outcome before applying it: Turns transcripts into summaries, actions, owners, and knowledge notes.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Create a source-backed outline and mark uncertainty. Approve no mutation unless the preview matches your intent.

You are done when
Meeting Transcript Processor is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 15

Use the Book / Report Summarizer Work Mode

Compresses long documents into useful notes, frameworks, and action points. This mode belongs to the Research & Knowledge family.

Beginner5–15 minResult · Book / Report Summarizer is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Research & Knowledge family, and select Book / Report Summarizer.

  2. Read the exact outcome before applying it: Compresses long documents into useful notes, frameworks, and action points.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Create a source-backed outline and mark uncertainty. Approve no mutation unless the preview matches your intent.

You are done when
Book / Report Summarizer is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 16

Use the Local Knowledge Vault Work Mode

Persistent local memory plus RAG-style retrieval for trusted on-device context. This mode belongs to the Research & Knowledge family.

Beginner5–15 minResult · Local Knowledge Vault is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Research & Knowledge family, and select Local Knowledge Vault.

  2. Read the exact outcome before applying it: Persistent local memory plus RAG-style retrieval for trusted on-device context.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Create a source-backed outline and mark uncertainty. Approve no mutation unless the preview matches your intent.

You are done when
Local Knowledge Vault is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 17

Use the Editor / Proofreader Work Mode

Polishes clarity, grammar, structure, and tone without flattening the voice. This mode belongs to the Writer & Content family.

Beginner5–15 minResult · Editor / Proofreader is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Writer & Content family, and select Editor / Proofreader.

  2. Read the exact outcome before applying it: Polishes clarity, grammar, structure, and tone without flattening the voice.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Improve this short draft while preserving its meaning. Approve no mutation unless the preview matches your intent.

You are done when
Editor / Proofreader is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 18

Use the Copywriter Work Mode

Creates persuasive product, landing page, and campaign copy. This mode belongs to the Writer & Content family.

Beginner5–15 minResult · Copywriter is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Writer & Content family, and select Copywriter.

  2. Read the exact outcome before applying it: Creates persuasive product, landing page, and campaign copy.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Improve this short draft while preserving its meaning. Approve no mutation unless the preview matches your intent.

You are done when
Copywriter is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 19

Use the Technical Writer Work Mode

Turns product and code context into precise documentation. This mode belongs to the Writer & Content family.

Beginner5–15 minResult · Technical Writer is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Writer & Content family, and select Technical Writer.

  2. Read the exact outcome before applying it: Turns product and code context into precise documentation.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Improve this short draft while preserving its meaning. Approve no mutation unless the preview matches your intent.

You are done when
Technical Writer is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 20

Use the Blog Post Creator Work Mode

Builds long-form posts from ideas, outlines, and source material. This mode belongs to the Writer & Content family.

Beginner5–15 minResult · Blog Post Creator is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Writer & Content family, and select Blog Post Creator.

  2. Read the exact outcome before applying it: Builds long-form posts from ideas, outlines, and source material.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Improve this short draft while preserving its meaning. Approve no mutation unless the preview matches your intent.

You are done when
Blog Post Creator is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 21

Use the Social Media Manager Work Mode

Repurposes ideas into short-form content, variants, and cadence plans. This mode belongs to the Writer & Content family.

Beginner5–15 minResult · Social Media Manager is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Writer & Content family, and select Social Media Manager.

  2. Read the exact outcome before applying it: Repurposes ideas into short-form content, variants, and cadence plans.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Improve this short draft while preserving its meaning. Approve no mutation unless the preview matches your intent.

You are done when
Social Media Manager is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 22

Use the Research & Writing Partner Work Mode

Combines source gathering, outlining, and writing into one loop. This mode belongs to the Writer & Content family.

Beginner5–15 minResult · Research & Writing Partner is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Writer & Content family, and select Research & Writing Partner.

  2. Read the exact outcome before applying it: Combines source gathering, outlining, and writing into one loop.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Improve this short draft while preserving its meaning. Approve no mutation unless the preview matches your intent.

You are done when
Research & Writing Partner is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 23

Use the Content Repurposer Work Mode

Turns one source into multiple formats, lengths, and channels. This mode belongs to the Writer & Content family.

Beginner5–15 minResult · Content Repurposer is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Writer & Content family, and select Content Repurposer.

  2. Read the exact outcome before applying it: Turns one source into multiple formats, lengths, and channels.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Improve this short draft while preserving its meaning. Approve no mutation unless the preview matches your intent.

You are done when
Content Repurposer is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 24

Use the Inbox Zero Commander Work Mode

Triages inbound messages into replies, follow-ups, and action queues. This mode belongs to the Email & Messages family.

Beginner5–15 minResult · Inbox Zero Commander is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Email & Messages family, and select Inbox Zero Commander.

  2. Read the exact outcome before applying it: Triages inbound messages into replies, follow-ups, and action queues.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Draft a concise reply but do not send it. Approve no mutation unless the preview matches your intent.

You are done when
Inbox Zero Commander is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 25

Use the Email Composer Work Mode

Writes polished emails from rough notes, context, or prior threads. This mode belongs to the Email & Messages family.

Beginner5–15 minResult · Email Composer is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Email & Messages family, and select Email Composer.

  2. Read the exact outcome before applying it: Writes polished emails from rough notes, context, or prior threads.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Draft a concise reply but do not send it. Approve no mutation unless the preview matches your intent.

You are done when
Email Composer is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 26

Use the Reply Assistant Work Mode

Produces concise replies tuned to context, urgency, and relationship. This mode belongs to the Email & Messages family.

Beginner5–15 minResult · Reply Assistant is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Email & Messages family, and select Reply Assistant.

  2. Read the exact outcome before applying it: Produces concise replies tuned to context, urgency, and relationship.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Draft a concise reply but do not send it. Approve no mutation unless the preview matches your intent.

You are done when
Reply Assistant is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 27

Use the Slack / Message Polisher Work Mode

Refines short-form messages for tone, clarity, and speed. This mode belongs to the Email & Messages family.

Beginner5–15 minResult · Slack / Message Polisher is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Email & Messages family, and select Slack / Message Polisher.

  2. Read the exact outcome before applying it: Refines short-form messages for tone, clarity, and speed.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Draft a concise reply but do not send it. Approve no mutation unless the preview matches your intent.

You are done when
Slack / Message Polisher is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 28

Use the Meeting Follow-up Work Mode

Turns notes and transcripts into follow-up emails and action recaps. This mode belongs to the Email & Messages family.

Beginner5–15 minResult · Meeting Follow-up is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Email & Messages family, and select Meeting Follow-up.

  2. Read the exact outcome before applying it: Turns notes and transcripts into follow-up emails and action recaps.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Draft a concise reply but do not send it. Approve no mutation unless the preview matches your intent.

You are done when
Meeting Follow-up is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 29

Use the Difficult Conversation Coach Work Mode

Helps de-escalate, soften, or strengthen messages for sensitive situations. This mode belongs to the Email & Messages family.

Beginner5–15 minResult · Difficult Conversation Coach is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Email & Messages family, and select Difficult Conversation Coach.

  2. Read the exact outcome before applying it: Helps de-escalate, soften, or strengthen messages for sensitive situations.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Draft a concise reply but do not send it. Approve no mutation unless the preview matches your intent.

You are done when
Difficult Conversation Coach is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 30

Use the Omnipresent Assistant Work Mode

A broad daily-use assistant for quick questions, notes, reminders, and odds and ends. This mode belongs to the Personal Assistant family.

Beginner5–15 minResult · Omnipresent Assistant is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Personal Assistant family, and select Omnipresent Assistant.

  2. Read the exact outcome before applying it: A broad daily-use assistant for quick questions, notes, reminders, and odds and ends.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Turn this request into three small next actions. Approve no mutation unless the preview matches your intent.

You are done when
Omnipresent Assistant is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 31

Use the Reminders Companion Work Mode

Tracks lightweight commitments, deadlines, and next actions. This mode belongs to the Personal Assistant family.

Beginner5–15 minResult · Reminders Companion is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Personal Assistant family, and select Reminders Companion.

  2. Read the exact outcome before applying it: Tracks lightweight commitments, deadlines, and next actions.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Turn this request into three small next actions. Approve no mutation unless the preview matches your intent.

You are done when
Reminders Companion is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 32

Use the Quick Research Work Mode

Answers fast practical questions without forcing a deep research workflow. This mode belongs to the Personal Assistant family.

Beginner5–15 minResult · Quick Research is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Personal Assistant family, and select Quick Research.

  2. Read the exact outcome before applying it: Answers fast practical questions without forcing a deep research workflow.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Turn this request into three small next actions. Approve no mutation unless the preview matches your intent.

You are done when
Quick Research is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 33

Use the Life Admin Work Mode

Organizes life logistics, forms, chores, and recurring admin tasks. This mode belongs to the Personal Assistant family.

Beginner5–15 minResult · Life Admin is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Personal Assistant family, and select Life Admin.

  2. Read the exact outcome before applying it: Organizes life logistics, forms, chores, and recurring admin tasks.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Turn this request into three small next actions. Approve no mutation unless the preview matches your intent.

You are done when
Life Admin is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 34

Use the Notes Copilot Work Mode

Turns rough notes into organized summaries and reusable memory. This mode belongs to the Personal Assistant family.

Beginner5–15 minResult · Notes Copilot is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Personal Assistant family, and select Notes Copilot.

  2. Read the exact outcome before applying it: Turns rough notes into organized summaries and reusable memory.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Turn this request into three small next actions. Approve no mutation unless the preview matches your intent.

You are done when
Notes Copilot is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 35

Use the Task Sidekick Work Mode

Breaks work into manageable next actions and light follow-up loops. This mode belongs to the Personal Assistant family.

Beginner5–15 minResult · Task Sidekick is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Personal Assistant family, and select Task Sidekick.

  2. Read the exact outcome before applying it: Breaks work into manageable next actions and light follow-up loops.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Turn this request into three small next actions. Approve no mutation unless the preview matches your intent.

You are done when
Task Sidekick is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 36

Use the Messaging Bridge Work Mode

Brings personal assistant behaviors into message-driven workflows. This mode belongs to the Personal Assistant family.

Beginner5–15 minResult · Messaging Bridge is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Personal Assistant family, and select Messaging Bridge.

  2. Read the exact outcome before applying it: Brings personal assistant behaviors into message-driven workflows.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Turn this request into three small next actions. Approve no mutation unless the preview matches your intent.

You are done when
Messaging Bridge is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 37

Use the Daily Briefing Butler Work Mode

Produces a concise operational briefing with priorities, blockers, and follow-ups. This mode belongs to the Chief of Staff family.

Beginner5–15 minResult · Daily Briefing Butler is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Chief of Staff family, and select Daily Briefing Butler.

  2. Read the exact outcome before applying it: Produces a concise operational briefing with priorities, blockers, and follow-ups.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Create a short briefing with priorities and blockers. Approve no mutation unless the preview matches your intent.

You are done when
Daily Briefing Butler is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 38

Use the Smart Scheduler Work Mode

Coordinates timing, sequencing, and follow-up pressure across commitments. This mode belongs to the Chief of Staff family.

Beginner5–15 minResult · Smart Scheduler is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Chief of Staff family, and select Smart Scheduler.

  2. Read the exact outcome before applying it: Coordinates timing, sequencing, and follow-up pressure across commitments.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Create a short briefing with priorities and blockers. Approve no mutation unless the preview matches your intent.

You are done when
Smart Scheduler is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 39

Use the Morning Briefing Work Mode

Starts the day with a short, decision-ready overview. This mode belongs to the Chief of Staff family.

Beginner5–15 minResult · Morning Briefing is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Chief of Staff family, and select Morning Briefing.

  2. Read the exact outcome before applying it: Starts the day with a short, decision-ready overview.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Create a short briefing with priorities and blockers. Approve no mutation unless the preview matches your intent.

You are done when
Morning Briefing is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 40

Use the Calendar-Aware Coordination Work Mode

Keeps meetings, deadlines, and commitments aligned with real capacity. This mode belongs to the Chief of Staff family.

Beginner5–15 minResult · Calendar-Aware Coordination is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Chief of Staff family, and select Calendar-Aware Coordination.

  2. Read the exact outcome before applying it: Keeps meetings, deadlines, and commitments aligned with real capacity.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Create a short briefing with priorities and blockers. Approve no mutation unless the preview matches your intent.

You are done when
Calendar-Aware Coordination is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 41

Use the Follow-up System Work Mode

Tracks open loops, nudges, and status changes across work streams. This mode belongs to the Chief of Staff family.

Beginner5–15 minResult · Follow-up System is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Chief of Staff family, and select Follow-up System.

  2. Read the exact outcome before applying it: Tracks open loops, nudges, and status changes across work streams.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Create a short briefing with priorities and blockers. Approve no mutation unless the preview matches your intent.

You are done when
Follow-up System is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 42

Use the Persistent Workflow Support Work Mode

Maintains context across ongoing work, priorities, and operational rhythms. This mode belongs to the Chief of Staff family.

Beginner5–15 minResult · Persistent Workflow Support is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Chief of Staff family, and select Persistent Workflow Support.

  2. Read the exact outcome before applying it: Maintains context across ongoing work, priorities, and operational rhythms.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Create a short briefing with priorities and blockers. Approve no mutation unless the preview matches your intent.

You are done when
Persistent Workflow Support is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 43

Use the Desktop Janitor Work Mode

Sorts clutter, cleanup opportunities, and low-risk local organization work. This mode belongs to the Mac Operator family.

Intermediate5–15 minResult · Desktop Janitor is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Mac Operator family, and select Desktop Janitor.

  2. Read the exact outcome before applying it: Sorts clutter, cleanup opportunities, and low-risk local organization work.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Propose a reversible plan; do not change files yet. Approve no mutation unless the preview matches your intent.

You are done when
Desktop Janitor is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 44

Use the Local File Workflows Work Mode

Runs file transforms, organization passes, and folder-centric work. This mode belongs to the Mac Operator family.

Intermediate5–15 minResult · Local File Workflows is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Mac Operator family, and select Local File Workflows.

  2. Read the exact outcome before applying it: Runs file transforms, organization passes, and folder-centric work.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Propose a reversible plan; do not change files yet. Approve no mutation unless the preview matches your intent.

You are done when
Local File Workflows is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 45

Use the Browser Tasks Work Mode

Handles browser-assisted tasks while keeping navigation and actions visible. This mode belongs to the Mac Operator family.

Intermediate5–15 minResult · Browser Tasks is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Mac Operator family, and select Browser Tasks.

  2. Read the exact outcome before applying it: Handles browser-assisted tasks while keeping navigation and actions visible.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Propose a reversible plan; do not change files yet. Approve no mutation unless the preview matches your intent.

You are done when
Browser Tasks is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 46

Use the OS-Level Assistance Work Mode

Helps with safe local system tasks and troubleshooting. This mode belongs to the Mac Operator family.

Intermediate5–15 minResult · OS-Level Assistance is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Mac Operator family, and select OS-Level Assistance.

  2. Read the exact outcome before applying it: Helps with safe local system tasks and troubleshooting.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Propose a reversible plan; do not change files yet. Approve no mutation unless the preview matches your intent.

You are done when
OS-Level Assistance is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 47

Use the Safe Desktop Automation Work Mode

Chains local actions carefully under approval gates. This mode belongs to the Mac Operator family.

Intermediate5–15 minResult · Safe Desktop Automation is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Mac Operator family, and select Safe Desktop Automation.

  2. Read the exact outcome before applying it: Chains local actions carefully under approval gates.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Propose a reversible plan; do not change files yet. Approve no mutation unless the preview matches your intent.

You are done when
Safe Desktop Automation is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 48

Use the Folder Organizer Work Mode

Builds consistent folder structure and tidy naming patterns. This mode belongs to the Mac Operator family.

Intermediate5–15 minResult · Folder Organizer is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Mac Operator family, and select Folder Organizer.

  2. Read the exact outcome before applying it: Builds consistent folder structure and tidy naming patterns.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Propose a reversible plan; do not change files yet. Approve no mutation unless the preview matches your intent.

You are done when
Folder Organizer is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 49

Use the AppleScript / Shell Helper Work Mode

Produces safe local scripts for repeatable Mac tasks. This mode belongs to the Mac Operator family.

Intermediate5–15 minResult · AppleScript / Shell Helper is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Mac Operator family, and select AppleScript / Shell Helper.

  2. Read the exact outcome before applying it: Produces safe local scripts for repeatable Mac tasks.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Propose a reversible plan; do not change files yet. Approve no mutation unless the preview matches your intent.

You are done when
AppleScript / Shell Helper is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 50

Use the Workflow Builder Work Mode

Chains existing tools and services into repeatable workflows. This mode belongs to the Automation Agent family.

Intermediate5–15 minResult · Workflow Builder is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Automation Agent family, and select Workflow Builder.

  2. Read the exact outcome before applying it: Chains existing tools and services into repeatable workflows.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Design a reviewable flow and identify every approval gate. Approve no mutation unless the preview matches your intent.

You are done when
Workflow Builder is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 51

Use the Project Scaffolder Work Mode

Bootstraps new projects, templates, and repeatable setup work. This mode belongs to the Automation Agent family.

Intermediate5–15 minResult · Project Scaffolder is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Automation Agent family, and select Project Scaffolder.

  2. Read the exact outcome before applying it: Bootstraps new projects, templates, and repeatable setup work.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Design a reviewable flow and identify every approval gate. Approve no mutation unless the preview matches your intent.

You are done when
Project Scaffolder is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 52

Use the Documentation Sync Work Mode

Keeps docs aligned with code and changing system behavior. This mode belongs to the Automation Agent family.

Intermediate5–15 minResult · Documentation Sync is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Automation Agent family, and select Documentation Sync.

  2. Read the exact outcome before applying it: Keeps docs aligned with code and changing system behavior.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Design a reviewable flow and identify every approval gate. Approve no mutation unless the preview matches your intent.

You are done when
Documentation Sync is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 53

Use the Multi-Step Tool Use Work Mode

Coordinates browser, files, memory, and runtime tools as one flow. This mode belongs to the Automation Agent family.

Intermediate5–15 minResult · Multi-Step Tool Use is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Automation Agent family, and select Multi-Step Tool Use.

  2. Read the exact outcome before applying it: Coordinates browser, files, memory, and runtime tools as one flow.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Design a reviewable flow and identify every approval gate. Approve no mutation unless the preview matches your intent.

You are done when
Multi-Step Tool Use is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 54

Use the Approval-Gated Autonomous Execution Work Mode

A stronger autonomy preset that still keeps impactful actions behind review. This mode belongs to the Automation Agent family.

Intermediate5–15 minResult · Approval-Gated Autonomous Execution is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Automation Agent family, and select Approval-Gated Autonomous Execution.

  2. Read the exact outcome before applying it: A stronger autonomy preset that still keeps impactful actions behind review.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Design a reviewable flow and identify every approval gate. Approve no mutation unless the preview matches your intent.

You are done when
Approval-Gated Autonomous Execution is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 55

Use the OpenClaw-Style Orchestrator Work Mode

Centers orchestration, agent handoffs, and reviewable execution plans. This mode belongs to the Automation Agent family.

Intermediate5–15 minResult · OpenClaw-Style Orchestrator is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Automation Agent family, and select OpenClaw-Style Orchestrator.

  2. Read the exact outcome before applying it: Centers orchestration, agent handoffs, and reviewable execution plans.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Design a reviewable flow and identify every approval gate. Approve no mutation unless the preview matches your intent.

You are done when
OpenClaw-Style Orchestrator is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 56

Use the CSV Analyst Work Mode

Explains, summarizes, and transforms CSV data with natural language. This mode belongs to the Data Analyst family.

Beginner5–15 minResult · CSV Analyst is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Data Analyst family, and select CSV Analyst.

  2. Read the exact outcome before applying it: Explains, summarizes, and transforms CSV data with natural language.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Describe the checks you would run before trusting this data. Approve no mutation unless the preview matches your intent.

You are done when
CSV Analyst is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 57

Use the Spreadsheet Q&A Work Mode

Answers plain-language questions about spreadsheet-style data. This mode belongs to the Data Analyst family.

Beginner5–15 minResult · Spreadsheet Q&A is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Data Analyst family, and select Spreadsheet Q&A.

  2. Read the exact outcome before applying it: Answers plain-language questions about spreadsheet-style data.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Describe the checks you would run before trusting this data. Approve no mutation unless the preview matches your intent.

You are done when
Spreadsheet Q&A is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 58

Use the Log Analyzer Work Mode

Finds patterns, spikes, failures, and next debugging paths in logs. This mode belongs to the Data Analyst family.

Beginner5–15 minResult · Log Analyzer is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Data Analyst family, and select Log Analyzer.

  2. Read the exact outcome before applying it: Finds patterns, spikes, failures, and next debugging paths in logs.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Describe the checks you would run before trusting this data. Approve no mutation unless the preview matches your intent.

You are done when
Log Analyzer is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 59

Use the JSON / XML Transformer Work Mode

Transforms structured data into new shapes, schemas, or human-readable outputs. This mode belongs to the Data Analyst family.

Beginner5–15 minResult · JSON / XML Transformer is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Data Analyst family, and select JSON / XML Transformer.

  2. Read the exact outcome before applying it: Transforms structured data into new shapes, schemas, or human-readable outputs.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Describe the checks you would run before trusting this data. Approve no mutation unless the preview matches your intent.

You are done when
JSON / XML Transformer is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 60

Use the SQL Query Builder Work Mode

Turns analysis questions into precise SQL and explains the shape of the answer. This mode belongs to the Data Analyst family.

Beginner5–15 minResult · SQL Query Builder is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Data Analyst family, and select SQL Query Builder.

  2. Read the exact outcome before applying it: Turns analysis questions into precise SQL and explains the shape of the answer.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Describe the checks you would run before trusting this data. Approve no mutation unless the preview matches your intent.

You are done when
SQL Query Builder is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 61

Use the Regex Generator Work Mode

Builds and explains patterns for parsing and filtering structured text. This mode belongs to the Data Analyst family.

Beginner5–15 minResult · Regex Generator is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Data Analyst family, and select Regex Generator.

  2. Read the exact outcome before applying it: Builds and explains patterns for parsing and filtering structured text.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Describe the checks you would run before trusting this data. Approve no mutation unless the preview matches your intent.

You are done when
Regex Generator is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 62

Use the Natural-Language Data Workflows Work Mode

Combines ingestion, transforms, QA, and output shaping into one assistant loop. This mode belongs to the Data Analyst family.

Beginner5–15 minResult · Natural-Language Data Workflows is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Data Analyst family, and select Natural-Language Data Workflows.

  2. Read the exact outcome before applying it: Combines ingestion, transforms, QA, and output shaping into one assistant loop.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Describe the checks you would run before trusting this data. Approve no mutation unless the preview matches your intent.

You are done when
Natural-Language Data Workflows is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 63

Use the OpenAI-Compatible Local API Server Work Mode

Turns the app into a clean local endpoint for tools, scripts, and external apps. This mode belongs to the Local AI Endpoint family.

Intermediate5–15 minResult · OpenAI-Compatible Local API Server is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Local AI Endpoint family, and select OpenAI-Compatible Local API Server.

  2. Read the exact outcome before applying it: Turns the app into a clean local endpoint for tools, scripts, and external apps.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Show the smallest harmless client test for this endpoint. Approve no mutation unless the preview matches your intent.

You are done when
OpenAI-Compatible Local API Server is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 64

Use the Local Inference Server Configuration Work Mode

Optimizes ports, defaults, and routing for local inference serving. This mode belongs to the Local AI Endpoint family.

Intermediate5–15 minResult · Local Inference Server Configuration is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Local AI Endpoint family, and select Local Inference Server Configuration.

  2. Read the exact outcome before applying it: Optimizes ports, defaults, and routing for local inference serving.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Show the smallest harmless client test for this endpoint. Approve no mutation unless the preview matches your intent.

You are done when
Local Inference Server Configuration is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 65

Use the Authentication / Port Setup Work Mode

Sets up safe access defaults, auth, and clear connection details. This mode belongs to the Local AI Endpoint family.

Intermediate5–15 minResult · Authentication / Port Setup is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Local AI Endpoint family, and select Authentication / Port Setup.

  2. Read the exact outcome before applying it: Sets up safe access defaults, auth, and clear connection details.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Show the smallest harmless client test for this endpoint. Approve no mutation unless the preview matches your intent.

You are done when
Authentication / Port Setup is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 66

Use the Quick-Start Client Snippets Work Mode

Generates ready-to-use snippets for scripts, apps, and developer tools. This mode belongs to the Local AI Endpoint family.

Intermediate5–15 minResult · Quick-Start Client Snippets is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Local AI Endpoint family, and select Quick-Start Client Snippets.

  2. Read the exact outcome before applying it: Generates ready-to-use snippets for scripts, apps, and developer tools.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Show the smallest harmless client test for this endpoint. Approve no mutation unless the preview matches your intent.

You are done when
Quick-Start Client Snippets is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.
Workflow 67

Use the Shortcuts / External App Integrations Work Mode

Preps the local endpoint for Shortcuts, automation tools, and external apps. This mode belongs to the Local AI Endpoint family.

Intermediate5–15 minResult · Shortcuts / External App Integrations is active and completes its starter test✓ PresetCatalog name, family, and summary verified
Before you start
  • At least one Ready model or provider route.
  • Review any capability, package, MCP, secret, or permission requirement shown by the mode.
  • Use disposable test content for the first run.
  1. Open Work Modes, find the Local AI Endpoint family, and select Shortcuts / External App Integrations.

  2. Read the exact outcome before applying it: Preps the local endpoint for Shortcuts, automation tools, and external apps.

  3. Open Requirements. Resolve blocking items only, review recommended items, and leave unrelated optional packages uninstalled.

  4. Click Start Here or Apply. Review the proposed execution mode, memory behavior, autonomy, tools, starter workspace, and automations before confirming.

  5. Run this harmless starter prompt: Show the smallest harmless client test for this endpoint. Approve no mutation unless the preview matches your intent.

You are done when
Shortcuts / External App Integrations is the active Work Mode, its health has no blocker, and the starter response matches the mode summary without an unexpected tool or external action.
If something looks wrong
A requirement is unavailableOpen the requirement action and resolve that single dependency, secret, MCP server, or package.
The mode changes too muchCancel, duplicate/customize the profile, lower autonomy, or disable starter automation before applying.
Sensitive work uses a cloud routeChange execution to Local Only and select a capable local model before testing real content.

Keep going