- Home
- Skills
- Technical Documentation
- Decision Workshop
Works with the AI Agents you already use
Decision Workshop
Turn a mixed architecture discussion with requirements, terminology, and open questions into accepted decisions, then route each outcome to one authoritative home: an ADR, spec, glossary, boundary, or investigation item.
$12
Decision Workshop
Example session with this skill installed
Close out this discussion using the current project authority files. We agreed to require audit logging. Feature flags are only a proposal. People are mixing “user” and “account,” and REST vs GraphQL remains unresolved. Separate accepted decisions from proposals and open questions; route each item to one authoritative artifact. Do not decide the API question or claim files were changed.
- Read your context and instructions
- Compiled the decision workshop
Illustrative worked example. The prompt supplies the facts below; no repository was inspected and no project files were changed.
ACCEPTED DECISIONS
- Require audit logging, stated as already agreed -> security behavior section of a Spec. The exact file is not supplied.
PROPOSALS
- Feature flags remain a proposal because the prompt does not say they were accepted. If rollout alternatives create a cross-cutting architecture choice, route it to an ADR after confirming the decision is in scope.
OPEN QUESTIONS
- REST vs GraphQL is unresolved. Route it to a focused decision pressure test after the team supplies query patterns and constraints. No recommendation is made here.
- “User” and “account” are being conflated. Route canonical meanings to the Glossary and revisit API naming after definitions are settled.
ARTIFACT ROUTING
- Audit logging requirement -> security section of a Spec.
- Feature flag rollout -> ADR candidate, not an accepted decision.
- User/account meanings -> Glossary.
- REST vs GraphQL -> investigation item for decision pressure testing.
DUPLICATION REMOVED / POINTERS ADDED
- None. No files were changed.
NEXT IMPLEMENTATION-READY INPUT
- Identify the authoritative Spec and Glossary, confirm the decision owner, and provide API query patterns and rollout constraints. Then update only authorized destinations.
Connects securely to your tools. The creator never sees your data.
About this skill
Decision Workshop closes out a mixed architecture discussion by separating accepted decisions, proposals, assumptions, terminology, and open questions, then routing each accepted outcome to one authoritative artifact such as an ADR, spec, glossary, boundary, or investigation item. Use it when project requirements, architecture choices, terminology, and unresolved questions have become tangled.
What it does
- Separates observed facts, proposed choices, assumptions, requirements, terminology, and unresolved questions. Repeated opinions do not become accepted decisions by repetition.
- Resolves only questions whose answers could change the choice. It routes a consequential unresolved choice to a focused decision pressure test and a terminology blocker to domain language governance.
- Gives each accepted outcome one authoritative destination: a glossary, ADR, spec, boundary or architecture rule, or investigation item. Secondary documents can point to that authority instead of duplicating it.
- Returns a structured closeout with accepted decisions, open questions, artifact routing, duplication or pointer notes, and the next implementation-ready input.
How it works
- Identify the decision scope and current authority files.
- Inventory facts, proposals, assumptions, requirements, terminology, and unresolved questions.
- Resolve decision-changing gaps or hand them to the appropriate focused workflow.
- Route each accepted outcome once, keeping unsupported items explicit.
- When writing is authorized, update only the selected authoritative artifacts. Without write access or authorization, return exact proposed updates and destinations.
Boundaries
Project governance docs, accepted ADRs, boundaries, glossaries, and approved specs outrank the skill. It does not manufacture consensus, treat repeated opinions as accepted authority, implement application code, or imply that commits, pushes, merges, installs, or deployments occurred.
How it differs from nearby workflows
Decision Pressure Test examines one consequential unsettled choice. Decision Workshop handles a mixed discussion with multiple decision states and routes only decision-changing uncertainty for deeper evaluation. Domain Language Governance handles terminology itself when word meaning blocks the choice; Decision Workshop identifies and routes that issue as part of the broader closeout.
Output
ACCEPTED DECISIONS: OPEN QUESTIONS: ARTIFACT ROUTING: DUPLICATION REMOVED / POINTERS ADDED: NEXT IMPLEMENTATION-READY INPUT:
The workflow depends on accessible project authority and enough evidence to distinguish accepted decisions from proposals. It cannot manufacture agreement or resolve a missing decision owner. It hands one consequential uncertainty to a focused decision pressure test and terminology blockers to domain language governance. Without authorized write access, it returns proposed changes rather than editing files. It does not implement application code or claim commits, pushes, deployments, installs, or validation that was not performed.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 17 days ago
Permissions
Read access to the conversation and current project authority/evidence files is required. Write Files is optional: use it only for explicitly authorized edits to selected authoritative artifacts; without it, return exact proposed updates and destinations. Terminal/Shell, Browser, Network Access, MCP, external accounts, and Environment Variables are not required. No commits, pushes, merges, installs, or deployments are implied.