Design Orchestration

1---
2name: design-orchestration
3description: >
4  Orchestrates design workflows by routing work through
5  brainstorming, multi-agent review, and execution readiness
6  in the correct order. Prevents premature implementation,
7  skipped validation, and unreviewed high-risk designs.
8---
9 
10# Design Orchestration (Meta-Skill)
11 
12## Purpose
13 
14Ensure that **ideas become designs**, **designs are reviewed**, and
15**only validated designs reach implementation**.
16 
17This skill does not generate designs.
18It **controls the flow between other skills**.
19 
20---
21 
22## Operating Model
23 
24This is a **routing and enforcement skill**, not a creative one.
25 
26It decides:
27- which skill must run next
28- whether escalation is required
29- whether execution is permitted
30 
31---
32 
33## Controlled Skills
34 
35This meta-skill coordinates the following:
36 
37- `brainstorming` — design generation
38- `multi-agent-brainstorming` — design validation
39- downstream implementation or planning skills
40 
41---
42 
43## Entry Conditions
44 
45Invoke this skill when:
46- a user proposes a new feature, system, or change
47- a design decision carries meaningful risk
48- correctness matters more than speed
49 
50---
51 
52## Routing Logic
53 
54### Step 1 — Brainstorming (Mandatory)
55 
56If no validated design exists:
57 
58- Invoke `brainstorming`
59- Require:
60  - Understanding Lock
61  - Initial Design
62  - Decision Log started
63 
64You may NOT proceed without these artifacts.
65 
66---
67 
68### Step 2 — Risk Assessment
69 
70After brainstorming completes, classify the design as:
71 
72- **Low risk**
73- **Moderate risk**
74- **High risk**
75 
76Use factors such as:
77- user impact
78- irreversibility
79- operational cost
80- complexity
81- uncertainty
82- novelty
83 
84---
85 
86### Step 3 — Conditional Escalation
87 
88- **Low risk**  
89  → Proceed to implementation planning
90 
91- **Moderate risk**  
92  → Recommend `multi-agent-brainstorming`
93 
94- **High risk**  
95  → REQUIRE `multi-agent-brainstorming`
96 
97Skipping escalation when required is prohibited.
98 
99---
100 
101### Step 4 — Multi-Agent Review (If Invoked)
102 
103If `multi-agent-brainstorming` is run:
104 
105Require:
106- completed Understanding Lock
107- current Design
108- Decision Log
109 
110Do NOT allow:
111- new ideation
112- scope expansion
113- reopening problem definition
114 
115Only critique, revision, and decision resolution are allowed.
116 
117---
118 
119### Step 5 — Execution Readiness Check
120 
121Before allowing implementation:
122 
123Confirm:
124- design is approved (single-agent or multi-agent)
125- Decision Log is complete
126- major assumptions are documented
127- known risks are acknowledged
128 
129If any condition fails:
130- block execution
131- return to the appropriate skill
132 
133---
134 
135## Enforcement Rules
136 
137- Do NOT allow implementation without a validated design
138- Do NOT allow skipping required review
139- Do NOT allow silent escalation or de-escalation
140- Do NOT merge design and implementation phases
141 
142---
143 
144## Exit Conditions
145 
146This meta-skill exits ONLY when:
147- the next step is explicitly identified, AND
148- all required prior steps are complete
149 
150Possible exits:
151- “Proceed to implementation planning”
152- “Run multi-agent-brainstorming”
153- “Return to brainstorming for clarification”
154- "If a reviewed design reports a final disposition of APPROVED, REVISE, or REJECT, you MUST route the workflow accordingly and state the chosen next step explicitly."
155---
156 
157## Design Philosophy
158 
159This skill exists to:
160- slow down the right decisions
161- speed up the right execution
162- prevent costly mistakes
163 
164Good systems fail early.
165Bad systems fail in production.
166 
167This meta-skill exists to enforce the former.
168