AI Delegation Brief Builder

Overview

Use this prompt-only skill when a user wants to delegate one concrete task to an AI assistant, agent, or model but the request is too vague to execute or verify. The skill turns the task into a compact delegation brief that explains the desired outcome, gives only useful context, sets boundaries, and defines how the user will judge the returned work.

The goal is not to create a universal prompt library. The goal is a single copy-paste brief for one task, with enough structure that the AI knows what to do and the user knows whether the result is good enough.

When to Use

Use this skill when the user asks to:

• write a better prompt for a specific AI task
• delegate a task to an AI assistant or agent
• clarify an AI work request before sending it
• turn notes into an AI task brief
• add acceptance criteria to an AI prompt
• define output format, constraints, and review checks for AI-generated work

Trigger keywords: AI task delegation brief, AI handoff, prompt for one task, acceptance criteria for AI output, task brief template, AI work request, delegate this to AI

Required Inputs

Ask only for details needed to build the brief:

• The task outcome the user wants
• The audience or end use for the output
• Available inputs or source material
• Deadline or urgency, if relevant
• Preferred output format, tone, length, or structure
• Important constraints, non-goals, examples, or forbidden actions
• The risk if the output is wrong or incomplete

If the user does not have all details, mark missing information as assumptions or follow-up questions instead of stalling.

Workflow

1. Name the delegated task. State the task in one sentence, including the desired outcome and where the output will be used.
2. Frame the success target. Capture audience, purpose, deadline, quality bar, and what good enough means.
3. Sort the context. Separate essential background from optional context so the final prompt stays concise.
4. List inputs and gaps. Identify source material, required files or excerpts, assumptions, and missing information.
5. Specify the output. Define format, sections, length, tone, style, examples to imitate, and examples to avoid.
6. Set boundaries. Add constraints, non-goals, privacy limits, forbidden actions, and decisions the AI must not make independently.
7. Write process instructions. Give the AI clear steps to follow before answering, including how to handle ambiguity.
8. Add acceptance criteria. Define observable checks the user can apply to the returned work.
9. Create the final brief. Produce a copy-paste prompt plus optional follow-up prompts for revision, verification, or expansion.

Output Format

Produce the delegation brief with these sections:

1. Delegation Snapshot
   • Task
   • Intended use
   • Audience
   • Deadline or urgency
   • Risk level
2. Context to Include
   • Essential background
   • Optional context
   • Source material or inputs
   • Assumptions and unknowns
3. Output Requirements
   • Format
   • Length or scope
   • Tone and style
   • Required sections
   • Examples to follow or avoid
4. Constraints and Boundaries
   • Must include
   • Must avoid
   • Non-goals
   • Privacy or confidentiality limits
   • Actions requiring human approval
5. Instructions for the AI
   • Step-by-step work plan
   • How to handle missing information
   • How to flag uncertainty
6. Acceptance Criteria
   • Checklist of observable pass conditions
   • Common failure modes to watch for
7. Copy-Paste Delegation Prompt
   • A polished prompt ready to send
8. Follow-Up Prompts
   • Clarify
   • Revise
   • Verify
   • Shorten or expand

Safety & Compliance

Explicit Boundaries

• No credential collection. Do not ask for passwords, API keys, recovery codes, account tokens, private keys, or full payment details.
• No unnecessary sensitive data. Encourage the user to redact private IDs, confidential documents, proprietary datasets, health details, legal details, financial account numbers, and third-party personal data unless a safe excerpt is essential.
• No autonomous external action. The brief must not instruct an AI to send messages, publish content, purchase items, contact people, delete data, change settings, or make binding decisions without explicit human review.
• No professional substitution. Legal, medical, financial, safety-critical, employment, academic, and regulatory tasks require qualified or authoritative review before reliance.
• No hidden assumptions. Missing facts should be labeled as assumptions, open questions, or placeholders.

Additional Safety Notes

• Keep the brief focused on one delegated task.
• Prefer placeholders such as [redacted source excerpt] over sensitive raw data.
• Add a human-review step before external use.
• If the task is high risk, tighten acceptance criteria and require source verification.

Acceptance Criteria

1. Produces a ready-to-send delegation brief for one concrete AI task.
2. Clearly states the trigger, goal, audience, end use, and quality bar.
3. Separates essential context from optional context.
4. Lists required inputs, missing information, and assumptions.
5. Defines output format, tone, length, sections, and examples.
6. Includes constraints, non-goals, privacy boundaries, and forbidden actions.
7. Converts the task into clear process instructions for the AI.
8. Adds acceptance criteria and a review checklist.
9. Includes follow-up prompts for clarifying, revising, or verifying the output.
10. Avoids credential requests, sensitive-data exposure, and autonomous external actions.

Example

User says: "I need an AI to draft a customer onboarding email sequence, but I keep getting generic results."

Skill response: Build a delegation brief with the campaign goal, audience, product context, source notes, tone, sequence length, constraints, non-goals, acceptance criteria, and a final copy-paste prompt the user can send to an AI assistant.