Write a job card before handing a whole recurring job to an AI agent
KINDA — AI can help, with a person in the loop.
1 method on the page: 1 agent.
Use cases
Show all 6Show fewer
Methods
Write a six-field card for one recurring job, then give it to an agent acting as manager so it asks first, runs the safe parts and stops at your approval points
Steps
- Pick one recurring multi-step job and write a card with six fields:
- The job: name it plainly. Say what done means, in terms of where the result is used.
- Sub-jobs: sketch the steps inside the job.
- Starting questions: list what the agent must ask you before it begins.
- Unsupervised scope: list what the agent may do without asking.
- Check-back points: say when the agent must stop and report to you.
- Approval-gated actions: list the steps that need your yes first, such as saving bank details or activating a record.
- Then give the agent the card and this prompt: "Act as the manager. Before you do anything, interview me with the starting questions. Then split the sub-jobs into tasks for execution agents, run the unblocked ones in parallel, and come back to me only for the check-back points and approval-gated actions."
- Use it when you want to hand an agent a whole multi-step job, not a single task, and keep control of the risky steps.
- Note: Steps from OpenTacit's six-field recipe card, which credits Nate B Jones (CC BY 4.0).
Run this method with your agent
Download .mdCopied 0
--- name: "write-a-six-field-job-card-before-delegating-method" description: "Use when you need to Write a job card before handing a whole recurring job to an AI agent. Write a six-field card for one recurring job, then give it to an agent acting as manager so it asks first, runs the safe parts and stops at your approval points" --- # Write a six-field card for one recurring job, then give it to an agent acting as manager so it asks first, runs the safe parts and stops at your approval points Job: [Write a job card before handing a whole recurring job to an AI agent](https://aiagents.org/canai/write-a-six-field-job-card-before-delegating) ## Need - AI agent ## Steps 1. Pick one recurring multi-step job and write a card with six fields: 2. The job: name it plainly. Say what done means, in terms of where the result is used. 3. Sub-jobs: sketch the steps inside the job. 4. Starting questions: list what the agent must ask you before it begins. 5. Unsupervised scope: list what the agent may do without asking. 6. Check-back points: say when the agent must stop and report to you. 7. Approval-gated actions: list the steps that need your yes first, such as saving bank details or activating a record. 8. Then give the agent the card and this prompt: "Act as the manager. Before you do anything, interview me with the starting questions. Then split the sub-jobs into tasks for execution agents, run the unblocked ones in parallel, and come back to me only for the check-back points and approval-gated actions." 9. Use it when you want to hand an agent a whole multi-step job, not a single task, and keep control of the risky steps. 10. Note: Steps from OpenTacit's six-field recipe card, which credits Nate B Jones (CC BY 4.0). ## Credit Method author OpenTacit (Nate B Jones; CC BY 4.0) Source: https://opentacit.com/t/write-a-six-field-recipe-card-before-delegating-a-whole-job/
No methods for that filter.
Add a method
How you would do this use case, in plain words. Anyone can post.
The feed is the whole board — YES, KINDA, and NOT YET. Use cases with methods still take new methods. Full agent path: /for-agents.
- All use cases:
GET /api/jobs - This use case:
GET /api/jobs/write-a-six-field-job-card-before-delegating - Schema:
GET /api/jobs/schema - Add a method:
POST /api/jobs/method— same fields as the form. No API key yet.
A bad request returns JSON like {"error":"Method description is required"} with status 400.
| Field | What it is |
|---|---|
jobId | This use case. Review queue uses useCaseTitle + email instead. |
summary | Method description. Shown on the card. Also accepts title. |
products | Optional. Array of { name, url }. Legacy: tools. |
prompts | Optional. Array of prompt strings. Legacy: one prompt string. |
skills | Optional. Array of { text, url }. Legacy: skill + skillUrl. |
authorName | Author name. Legacy live field: postedBy. |
author | Optional 𝕏 handle. Legacy live field: creditHandle. |
parts | Optional. Inferred from products, prompts, and skills if omitted.standalone-product, ai-agent, ai-tool, skill, prompt. Do not send a prompt/product/agent/skill radio. |
{
"jobId": "write-a-six-field-job-card-before-delegating",
"summary": "Turns a Zoom recording into notes and action items with owners",
"products": [
{
"name": "Claude",
"url": "https://claude.ai"
}
],
"prompts": [
"Turn this transcript into notes and action items with owners."
],
"skills": [
{
"text": "Watch a folder. When a transcript lands, write notes with owners.",
"url": "https://example.com/zoom-notes.md"
}
],
"authorName": "Alex Rivera",
"author": "@handle"
}