Engagement Context
Why the engagement exists, what precedes it, and the surrounding organizational conditions.
Free portable AI skill
STACK Engagement Designer
A purposeful engagement design that connects Situation, Task, Action, Consequence, and Knowledge, with participatory Scan-Focus-Act cycles, appropriate AI support, human decision rights, and follow-through.
What you need to provide.
The skill names unknowns instead of inventing missing facts.
Participants And Stakeholders
Who will participate, who is affected, their relevant knowledge, and accessibility needs.
Task And Decision Rights
The measurable task, decision to support, decision owner, and what participants can influence.
Logistics And Constraints
Time, setting, group size, technology, data, budget, policy, privacy, and facilitation capacity.
Desired Consequences And Knowledge
Intended outputs, actions, learning, and evidence to preserve after the engagement.
The six-pass workflow.
Each pass leaves an artifact that a human owner can inspect, revise, or stop.
Situation
Build an evidence-based view of context, history, participants, constraints, stakeholder needs, and uncertainties.
Task
Define a specific engagement objective, decision boundary, tangible output, success evidence, and accountable decision owner.
Action
Design the participant experience as one or more Scan-Focus-Act cycles, naming activities, prompts, roles, timeboxes, AI assistance, human synthesis, and fallback methods.
Consequence
Examine intended and unintended outcomes, commitments, risks, dependencies, and what must happen immediately after the engagement.
Knowledge
Specify how inputs, decisions, dissent, white-space ideas, action owners, and lessons will be captured, validated, shared, and reused.
Stress-test and hand off
Test participation, accessibility, privacy, tool failure, evidence quality, decision authority, and follow-through before producing the run sheet.
What the skill returns.
Outputs are decision-support artifacts, not autonomous decisions.
Stack Design Brief
Situation, Task, Action, Consequence, and Knowledge in one concise design view.
Facilitator Run Sheet
A timed sequence of activities, prompts, roles, transitions, and artifacts.
Scan Focus Act Cycles
For each cycle, the input to scan, focusing criteria, action or test, and feedback into the next cycle.
Human Ai Role Map
Where AI assists and where humans facilitate, validate, decide, own, or intervene.
Consequence And Knowledge Plan
Commitments, follow-through, knowledge capture, validation, access, and reuse.
Guardrails before use.
Keep these conditions visible when adapting the prompt to your policies and risk profile.
- Do not use AI synthesis as a substitute for participant voice, facilitator judgment, or an authorized decision.
- Do not infer consensus from silence, participation counts, sentiment analysis, or model summaries.
- Disclose material AI use to participants and collect consent where required.
- Protect personal, confidential, and sensitive contributions; define retention and access before capture.
- Provide an accessible non-AI fallback for critical activities and technology failure.
- Do not design beyond the actual decision rights, time, staffing, or evidence available.
- Name an owner and checkpoint for every consequential commitment.
The complete copy-paste prompt.
Replace bracketed fields with permitted context, then use an approved assistant.
STACK Engagement Designer
Act as a UWT STACK Engagement Designer. Design one purposeful engagement using Situation, Task, Action, Consequence, and Knowledge. Within Action, use one or more rapid Scan-Focus-Act cycles. AI may assist with research, synthesis, option generation, pattern detection, or documentation, but humans must retain facilitation, validation, accountability, and consequential decision rights. Do not invent participant needs, evidence, consensus, or tool capabilities. Context: - Engagement purpose and background: [PASTE] - Participants and affected stakeholders: [PASTE] - Decision or task: [PASTE] - Decision owner and participant influence: [PASTE] - Desired outputs and consequences: [PASTE] - Time, setting, group size, accessibility needs, and constraints: [PASTE] - Available data, tools, facilitators, and source material: [PASTE] - Privacy, policy, or consent requirements: [PASTE] Create the design in six passes: 1. Situation: summarize the evidence-based context, relevant history, participant knowledge, stakeholder needs, constraints, and unknowns. 2. Task: write a specific measurable task; identify the tangible artifact or decision; define success evidence; state what is outside scope. 3. Action: create a timed participant journey built from Scan-Focus-Act cycles. For every block, specify purpose, participant activity, prompt, facilitator move, AI contribution, human validation, artifact, and timebox. In Scan, widen and synthesize the evidence or option space. In Focus, name criteria and narrow deliberately while preserving dissent and white-space ideas. In Act, test, decide, prototype, commit, or move the work forward. 4. Consequence: list intended and possible unintended outcomes, immediate commitments, owners, dependencies, and the first post-engagement checkpoint. 5. Knowledge: define how source inputs, decisions, dissent, ideas, actions, and lessons will be captured, validated, stored, accessed, and reused. 6. Stress-test: check decision authority, participation equity, accessibility, privacy, evidence quality, tool failure, bias, and follow-through. Offer a non-AI fallback for every critical AI-supported step. Output: STACK design brief; timed facilitator run sheet; Scan-Focus-Act cycle table; human-AI role map; artifact list; technology and fallback plan; consequence and knowledge plan; unresolved questions. Do not call a model summary consensus unless participants and the authorized decision owner validate it.
Packaging boundary: The four items below are approved UWT prompt-based companion guides. They are not official Claude skills, OpenAI skills, GPTs, plugins, or packaged capabilities from any AI vendor.
Concepts behind the skill.
Every workflow links back to specific canonical manuscript sections.
Chapter 3 - Implementing UWT
Situation, Task, Action, Consequence, Knowledge, Scan-Focus-Act
Pages 93-99
Chapter 3 - Implementing UWT
implementation prompts, STACK questions, AI-assisted engagement design
Pages 100-102
Chapter 6 - The Rise of the Cognitor
process choreography, AI tool orchestration, rapid iteration
Pages 144-150