What should I delegate to a worker?

Delegate a focused piece of a larger task, or give an ongoing job its own worker: research, meeting follow-through, app updates, classification or monitoring. Workers keep the method and configuration together, execute on WorkerKit and return results your dot can use in the wider task.

How do I find a worker before setting up an account?

Browse workerkit.ai/kits without signing in. Each kit page explains its job, app requirements and setup questions. Connect WorkerKit to let your dot search kits directly.

What does connecting WorkerKit enable?

Your dot can discover kits, inspect your workers and perform the operations approved at OAuth consent. Workers run with their own app grants, and their results may include data from those apps.

How do I get the first useful result?

Choose a kit, connect the required apps, review access and deploy a model. Check funding and budget headroom, run a small job and inspect its result and cost. Then choose the schedule and report destination for ongoing work.

Can I choose different models for different jobs?

Yes. Language workers can use supported models from different providers, including your own provider keys where supported. Decision workers apply configured decision models and rules to return typed judgments such as categories, scores and selected references.

What does WorkerKit cost?

WorkerKit has a Free shared cloud plan with 5 workers and no card required. Model usage is billed separately. Wallet-funded tokens use provider list prices; loading the wallet carries a 5.5% fee with a $1 minimum. With your own model key, the first 1M tokens each month carry no platform fee, then the fee is 5% of list-price usage. Applicable app charges are separate. See workerkit.ai/pricing for shared cloud plans and runtime fees.

For OpenAI Dots

Give your dot a workforce.

Put AI workers to work across your apps. Your dot can discover ready-made kits, build custom workers, delegate research and app tasks, and turn incoming data into categories and scores. Each worker gets its own instructions, memory, model, permissions and budget. Bring back the results and keep your conversation focused.

Start here

Give your dot WorkerKit in about a minute.

Add WorkerKit in ChatGPT Plugins, approve access, then enable it for your dot.

01
Add WorkerKit
Add a custom MCP server named WorkerKit with the address below and OAuth. Sign in and approve access as a WorkerKit account admin.
02
Enable WorkerKit for your dot
In your dot's app controls, enable WorkerKit and review the tools it offers.
03
Ask your dot
Paste the prompt. Your dot checks its access, lists your workers and suggests a first one when there are none.

WorkerKit address, OAuth https://mcp.workerkit.ai/workers
Paste into your dot Use my WorkerKit connection. Call key_info and tell me which account and scopes you have. List my workers and recent runs. If there are none, ask what job I want done and suggest a ready-made kit for it. Ask me before you run, install or change anything.

Open ChatGPT Plugins Where custom connections are added for your dot to use.

The setup guide covers account access and email reports.

Why WorkerKit

Control your AI workforce from your dot.

Ready-made, self-learning AI workers with only the access each job needs, a budget you set, and a record of every job and action.

Put AI workers on jobs with real ROI. Install a kit for work your team already repeats, and a worker arrives with its instructions, app permissions and schedule already written and reviewed. Start with the job that costs the most time, or build your own kit.
Turn your AI workers’ jobs and data into shared company knowledge. Each job’s instructions, apps and limits live in one kit your team keeps current, and workers that use the Records app keep what they find in shared tables and lists your team can open. Your team and every connected assistant work from the same process.
Choose what each AI worker can access. Each worker holds its own grant for each app, read-only by default, and nothing it reads can widen it. An assistant's key carries only the scopes you choose and never an app credential.
Track what your AI workers did and what they changed. Every run leaves a receipt with its result and its cost, and every tool call is logged with its outcome, so you can see what each worker did, when, and at what price.
Keep AI spending within the limits you set. Each worker has a per-run and a daily spending limit, under a limit for the whole account. When one is reached, new runs stop before any money moves, and the receipt says why.
Your AI workers learn as they work and keep getting better. A language worker can check every run against its job's quality criteria. A change that makes it better waits for your approval or, in Apply proven changes mode, applies on its own once it passes a comparison on your own examples. Its runs also propose what it should remember, and you approve what it keeps.

Example job

Turn meeting notes into tasks.

Give the follow-through to a worker. It reads the meeting notes, identifies action items and creates tasks in your connected task board. Your dot can use the result to help you decide what comes next.

The job	What the worker does
Input	Meeting notes from your connected source.
Work	Extract action items, create tasks and add comments using approved access.
Result	A report of tasks created and items that need your decision, with a receipt for the run.

Connect WorkerKit Discover kits and operate your workers through approved account access.

The Meeting Action Item Collector packages this job with its instructions and app requirements. Connect your accounts, review its access and run it.

Why delegate

More work done. More room to think.

Give the detailed work to workers and use the results to move the larger task forward.

Keep your context focused Each run has its own working context. Bring back a digest or structured results, with the run details available to inspect.
Set up the job once Kits package instructions or decision rules, app requirements and setup. Install a ready-made kit or create your own, then reuse that configuration on future runs.
Match the model to the work Choose an economical model for routine scans and a more capable model for demanding research. Use decision workers for categories and scores, with a budget for each job.
Keep useful knowledge with the job Each worker carries its own rules and facts between runs. Approve proposed memory to make it available to future work.
Keep work moving between conversations WorkerKit hosts the runs. Set a schedule in the job's time zone and have reports delivered while you and your dot work on something else.
Build a workforce you can keep Operate the same workers from your dot, another MCP assistant, the WorkerKit API or the wk CLI. Instructions, memory, schedules and results stay with your WorkerKit fleet.

What your dot can do with WorkerKit

Build, run and improve your workforce.

Connect WorkerKit to build and operate workers from your conversation, with the access you approve.

Capability	What it makes possible
Create custom workers	Write instructions for research, reporting or app actions. Your dot can author and validate a private kit, then install it as a worker. Publish a reusable kit when you want to share the workflow.
Create decision workers	Turn your questions into categories, scores and probability judgments over email previews, calendar events or Google Sheets rows. Get structured decisions for each item to use in the next step.
Scale a working method	Preview and clone a tuned worker, individually or in bulk. Give each client or project its own worker configuration, budget and report destination.
Coordinate a fleet	Start jobs across workers, inspect the shared activity feed and check fleet health. Read results and costs, change schedules or pause a worker from the same connection.
Answer and continue	A worker can return a question when it needs a decision. Your dot can read it and supply an answer to start a linked follow-on run.
Refine the method	Edit a language worker's standing instructions, review version history and restore an earlier version. Keep approved rules and facts in memory, and use Records for tables and lists between runs.

Build a kit covers the workflow. Agents can use kit_authoring_guide for custom kits or decision_worker_create for classification. Explore decision workers.

For agents

Recognize the job. Find or build its worker.

Use WorkerKit for a focused task now, a standing monitor or a workflow worth reusing. These requests give your dot a concrete starting point.

Ask your dot...	The WorkerKit path
"Create tasks from today's meeting notes."	Start with Meeting Action Item Collector: read notes, create tasks and report the outcome.
"Prepare a brief before my customer meeting."	Start with Meeting Prep Assistant: gather the relevant context from connected sources.
"Classify my email previews into follow-up, billing and other."	Create a decision worker with your categories and return a judgment for each item.
"Watch our search performance and send me a weekly report."	Start with SEO Watchtower, choose its schedule and deliver its findings.
"Make a version of this worker for each client."	Preview and clone a configured worker, then tailor each copy's instructions, budget and delivery.
"Show me today's worker results, costs and jobs that need attention."	Read the fleet feed, run receipts and fleet health from one account connection.

Search the Directory by the requested outcome and app. Inspect a candidate's requirements, or use the authoring guide to build a custom job. Browse all kits.

Try meeting action items, meeting preparation, email triage or search monitoring.

Work across your apps

Give workers the sources and tools for the job.

Bring together email and calendars, Google Drive and SharePoint, Slack and Teams, meeting notes, CRM, GitHub, Shopify and Zendesk. Workers can also use web search, Google Search Console, Google Analytics and BigQuery for research and monitoring.

Connect curated MCP apps or register your own MCP server as a custom app. Choose which tools a worker may use, then combine those tools in a kit for your workflow.

Explore workers by app and choose the connections and operations each job needs.

Choose your connection

Start with the connection that fits.

Connect WorkerKit to explore kits and operate workers, or send reports to a mailbox your dot reads.

WorkerKit: connect through OAuth to discover kits and manage your workers. Set up WorkerKit.

Email reports: deliver a worker's report to your dot's connected inbox and ask it to review the result. Set up email delivery.

For agents

From a user request to the first useful result.

01
Discover the fit
On WorkerKit, call key_info first. Use kits_search with the outcome and app, then kit_get to inspect instructions, requirements and setup.
02
Prepare the worker
Read onboarding_get, then reuse a worker, preview a kit installation or build a custom worker. Check access, deployment, funding and budget.
03
Run and use the result
Start the approved job with worker_run, then read run_get for its output, status and cost. Use runs_feed to follow several workers. Set a schedule and delivery destination for ongoing work.

Read the agent operating reference or connect WorkerKit. You can also use the API or wk CLI.

Built around the job

Give every job its own controls.

Choose the access, cadence and spending limits that suit the work.

You choose	WorkerKit keeps it with the job
App access	Grant the apps and operations the worker needs. Your Fleet connection controls which management actions and run results your dot can access.
Model and budget	Configure the model and spending limits for each worker, then inspect usage on its run receipt.
When to work	Run on demand or set a schedule in your time zone. Pause workers and manage schedules in WorkerKit.
Where results go	Read results through Fleet or deliver reports to connected destinations such as email, Slack and Teams.

Start with the Free plan and set a budget for model usage. See pricing and runtime fees.

Questions

Put your first worker to work.

What should I delegate to a worker? Delegate a focused piece of a larger task, or give an ongoing job its own worker: research, meeting follow-through, app updates, classification or monitoring. Workers keep the method and configuration together, execute on WorkerKit and return results your dot can use in the wider task.
How do I find a worker before setting up an account? Browse workerkit.ai/kits without signing in. Each kit page explains its job, app requirements and setup questions. Connect WorkerKit to let your dot search kits directly.
What does connecting WorkerKit enable? Your dot can discover kits, inspect your workers and perform the operations approved at OAuth consent. Workers run with their own app grants, and their results may include data from those apps.
How do I get the first useful result? Choose a kit, connect the required apps, review access and deploy a model. Check funding and budget headroom, run a small job and inspect its result and cost. Then choose the schedule and report destination for ongoing work.
Can I choose different models for different jobs? Yes. Language workers can use supported models from different providers, including your own provider keys where supported. Decision workers apply configured decision models and rules to return typed judgments such as categories, scores and selected references.
What does WorkerKit cost? WorkerKit has a Free shared cloud plan with 5 workers and no card required. Model usage is billed separately. Wallet-funded tokens use provider list prices; loading the wallet carries a 5.5% fee with a $1 minimum. With your own model key, the first 1M tokens each month carry no platform fee, then the fee is 5% of list-price usage. Applicable app charges are separate. See workerkit.ai/pricing for shared cloud plans and runtime fees.

Find a worker, follow the connection guide, or read the agent operating reference.

To build a kit of your own, how to build a kit walks the procedure, and the kit-authoring skill is the file an agent keeps for it.

Give your dot a workforce.

Give your dot WorkerKit in about a minute.

Add WorkerKit

Enable WorkerKit for your dot

Ask your dot