Agent-Skills.md

Agent Skills: Workshop Designer

Design a half-day or full-day strategy workshop that ends with a committed decision and named owners — not a "great discussion" with no artifact. Outputs an hour-by-hour agenda, breakout prompts with expected outputs, pre-reads, facilitator playbook for when things go sideways, and the post-workshop artifact that justifies the day. Use for offsites, leadership alignment sessions, kickoff workshops, planning meetings >2 hours.

UncategorizedID: sruthir28/enterprise-ai-skills/workshop-designer

Repository

sruthir28/enterprise-ai-skills
sruthir28License: MIT
10621

Install this agent skill to your local

pnpm dlx add-skill https://github.com/sruthir28/enterprise-ai-skills/tree/HEAD/workshop-designer

Skill Files

Browse the full folder contents for workshop-designer.

Download Skill

Loading file tree…

workshop-designer/SKILL.md

Skill Metadata

Name
workshop-designer
Description
Design a half-day or full-day strategy workshop that ends with a committed decision and named owners — not a "great discussion" with no artifact. Outputs an hour-by-hour agenda, breakout prompts with expected outputs, pre-reads, facilitator playbook for when things go sideways, and the post-workshop artifact that justifies the day. Use for offsites, leadership alignment sessions, kickoff workshops, planning meetings >2 hours.

Workshop Designer

Most workshops fail the same way: a lot of conversation, an emergent "alignment" that no one writes down, and a calendar invite for a follow-up workshop. This skill designs sessions that end with a committed artifact and named owners, because that's the only thing that justifies pulling people out of their week.

The test for any workshop: what artifact ships out of this room, and who owns the next step? If you can't answer, don't run the workshop.

Design choices

  • One outcome, not three. A workshop with three goals achieves zero. Force-rank to one decision or one artifact.
  • Diverge then converge, never both at once. Brainstorm sessions and decision sessions are different brain modes. Mixing them creates "discussion" that converges on nothing.
  • Breakouts < 4 people. Above 4, one person dominates. Report-backs < 5 min each, or you lose the room.
  • Energy curve matters. Heavy thinking morning, application + decisions afternoon. Never schedule a decision moment right after lunch.
  • The artifact is the deliverable, not the meeting. Pre-build the template; the workshop fills it in. Leaving "we'll write it up later" is how alignment evaporates by Tuesday.
  • Facilitator plays referee, not participant. If you have a stake in the decision, you can't facilitate. Pick one or the other.

Inputs needed (interview if missing)

  1. Workshop goal — one sentence. "End with an agreed Q3 OKR set" / "End with a decision on whether to build, buy, or partner" / "End with a 90-day plan owned by named leads."
  2. Attendees. Names + roles. Cap at 12 for decision workshops, 20 for divergent (brainstorm) workshops.
  3. Duration. Half-day (4 hrs incl. break), full-day (6–7 hrs incl. lunch + breaks). Multi-day = different format.
  4. What's already decided going in. Constraints, locked-in choices. Surface these upfront so people don't re-litigate them.
  5. What's open. The actual decision space.
  6. What the artifact is. The thing that leaves the room. (Strategy on a page, OKRs, decision memo, prioritized roadmap, etc.)
  7. Who facilitates. Internal (cheaper, has context) or external (neutral, harder to derail).

Output format

1. Workshop charter (1 page, pre-read)

  • The one outcome
  • Decisions in scope vs. out of scope
  • Attendee list with role in the workshop (decider / contributor / observer)
  • Required pre-work

2. Hour-by-hour agenda

| Time | Block | Mode | Output | Owner | |------|-------|------|--------|-------| | 9:00 | Frame + outcome | Plenary | Shared understanding of one outcome | Facilitator | | 9:15 | Constraints recap | Plenary | Locked items on the wall, no debate | Sponsor | | 9:30 | Divergent: option generation | Breakout (3s) | 5–8 options per group | Breakouts | | 10:30 | Report-back + cluster | Plenary | One synthesized list, ~12 options | Facilitator | | 11:00 | Break | — | — | — | | 11:15 | Convergent: rank options | Plenary, dot-vote | Top 3 options | All | | 12:00 | Lunch | — | — | — | | 1:00 | Deep-dive top option | Breakout (5–6) | Filled-in decision template | Sub-team | | 2:30 | Plenary review + decide | Plenary | Go / no-go + owner | All | | 3:00 | Wrap + commitments | Plenary | Named owners, dates, next step | Facilitator | | 3:30 | Done | — | Filled artifact in hand | — |

3. Breakout playbook

For each breakout: the prompt (one question), expected output (concrete deliverable), time (strict), roles (timekeeper, scribe, reporter).

4. Facilitator's "what if" notes

The 5 things that go wrong and what to do:

  • "One exec dominates" → name it, rotate to round-robin
  • "Group can't converge in the convergent block" → cut to 2 options, force a coin-flip-or-rationale framing
  • "New constraint surfaces mid-day" → park it on a 'parking lot' wall, re-decide if it changes the call
  • "Sponsor changes the goal mid-workshop" → pause, get the new outcome in writing, ask the room if it still works
  • "Pre-reads weren't read" → 10 min silent reading of the 1-page summary, skip the original frame block

5. The artifact template

The thing the workshop fills in. Pre-built. Examples:

  • Strategy on a page (Aspiration / Where to play / How to win / Capabilities / Mgmt systems)
  • Q3 OKR set (3 objectives, 3 KRs each, owners)
  • Decision memo (Context / Options / Recommendation / Risks / Ask)

6. Post-workshop send (24-hour SLA)

What goes out within 24 hrs: the artifact + the commitments table (who owns what by when). Drafted before the workshop, populated during.

Process

  1. Lock the one outcome. Push back on multi-outcome workshops. Run two days or one workshop and a follow-up.
  2. Design the artifact first. Reverse-engineer the agenda from what the artifact needs.
  3. Plot the energy curve. Heavy thinking 10am–12pm, decision 2pm–3pm, never after lunch directly.
  4. Mode-tag every block: divergent / convergent / plenary / breakout. If two adjacent blocks are the same mode for more than 90 min, restructure.
  5. Write the facilitator script. Not the slides — the verbatim framing for each transition. "We just spent an hour generating options. Now we're going to converge to three. The criterion is X. Each person gets 3 dots."
  6. Pre-build the "what if" playbook. Workshops go sideways for the same 5 reasons.
  7. Calendar the post-send before the workshop. 24-hr SLA for the artifact. If you don't pre-commit, it slips.

Worked example

Workshop: Align the 12-person leadership team on Q3 strategic priorities. One outcome: Leave with a Q3 strategy-on-a-page signed off by all 12. Duration: Full day, 9:00–3:30. Constraints (decided coming in): Headcount frozen; AI investment ringfenced; no new product lines. Open: Where to invest the 30% engineering capacity that's currently uncommitted.

Pre-reads (sent 5 days before)

  • Q2 retrospective: top 3 wins, top 3 misses (1 page)
  • 6 candidate investment areas with 1-paragraph cases each (3 pages)
  • Strategy-on-a-page template, pre-filled with locked items (1 page)

Agenda

| Time | Block | Mode | Output | |------|-------|------|--------| | 9:00 | Frame: today we leave with the strategy-on-a-page filled. The one open question: where do we point the uncommitted 30% capacity? | Plenary | Shared frame | | 9:15 | Q2 reflect: 1 win + 1 miss from each leader (1 min each, no discussion) | Plenary | Public reflection, surfaces what's on people's minds | | 9:45 | Breakouts of 3: rank the 6 candidate investments + propose 1 outside the pre-read | Breakout | 4 ranked lists | | 10:30 | Report-back + cluster | Plenary | One unified list, ~7 options | | 11:00 | Break | | | | 11:15 | Silent dot-voting: each leader gets 5 dots, places on the 7 options | Convergent | Top 3 emerge | | 11:45 | Devil's advocate: 1 person assigned to argue against each of the top 3 | Plenary | Stress-tested top 3 | | 12:15 | Lunch | | | | 1:15 | Breakouts of 4: each group takes one of the top 3 and fills in a 1-page "what this looks like in Q3" (investment, expected outcome, risks) | Breakout | 3 filled 1-pagers | | 2:30 | Plenary: choose 1 (or up to 2 if separable) | Convergent | Decision | | 2:45 | Fill in the strategy-on-a-page live on screen | Plenary | Artifact complete | | 3:15 | Commitments: named owner for each Q3 priority, kickoff date | Plenary | Owners table | | 3:30 | Wrap | | Artifact + commitments in hand |

Facilitator's "what if"

  • Most likely failure: group can't converge to 1 in the 2:30 block. Pre-decide: if no 70% consensus by 2:45, sponsor (CEO) gets the tie-breaker. Communicate this upfront so it's not a surprise.
  • Second risk: Q2 retrospective opens old wounds. Time-box hard, no rebuttals, facilitator interrupts if anyone goes >90 sec.
  • Third risk: someone re-litigates the headcount freeze. Point at the wall ("locked items"), defer to a different forum.

Post-workshop send (drafted before, populated during)

Email within 24 hours, subject: "Q3 strategy: locked. What you own."

  • Strategy on a page (attached PDF, photo of physical version on the wall)
  • Commitments table: each leader's Q3 priority owner-name + first milestone date
  • Next sync: 2-week check-in with this same group

Common mistakes to avoid

  • Multi-outcome workshops. "Alignment + brainstorm + decisions" = none of the three. Pick one.
  • No artifact template. "We'll write it up" = it doesn't get written up. Pre-build.
  • Schedule fills, work doesn't. Tight timeboxes feel good but don't generate output. Build in 15 min of slippage per major block.
  • Mixing divergent and convergent. "Let's brainstorm options and pick the best one in 45 min" produces neither good options nor a defensible pick.
  • Decision moment right after lunch. Energy nadir. Schedule application work there, not commitment.
  • >4 in a breakout. Loud voice dominates. Always split into 3s or 4s.
  • No "what if" playbook. Workshops sideways predictably. Have the moves ready.
  • The sponsor facilitates. They have a stake — they'll bias the room. Pick a neutral facilitator or rotate someone external in.

When to use

  • Leadership offsites (half-day or full-day)
  • Quarterly planning sessions
  • Strategy resets after a major change
  • Kickoff workshops for cross-functional initiatives
  • M&A integration alignment
  • Any working session > 2 hours

When NOT to use

  • Recurring meetings (use a different format)
  • Status updates (workshop is too much; use an async memo)
  • 1:1s, small group decisions <5 people (use Meeting Prep Kit)
  • Trainings (different design — facilitation vs. learning are different jobs)

Pairs well with

  • Meeting Prep Kit — pre-workshop 1:1s with key attendees to map stances; reduces in-room surprises
  • Stakeholder Map — map the room before designing the agenda; the breakouts you design depend on who's in the room
  • Decision Memo Builder — the artifact that leaves a decision workshop is often a decision memo
  • Storyline Builder — pre-reads structured as a storyline land better than narrative pre-reads