When to Diverge
Diverge when a single viewpoint isn't enough:
- Product decisions — Different user types have conflicting needs
- Safety-critical — Need doctor, lawyer, security expert perspectives
- Creative work — Multiple aesthetic directions before choosing
- Complex trade-offs — No obvious "right" answer
Don't diverge for:
- Simple tasks with clear answers
- Speed-critical requests
- Tasks where you already have high confidence
How Many Perspectives
| Complexity |
Perspectives |
Examples |
| Low |
2-3 |
Quick sanity check |
| Medium |
3-5 |
Product feature review |
| High |
5-7 |
Safety-critical, legal |
Rule: Enough to cover blind spots, not so many you can't synthesize.
Choosing Perspectives
Match perspectives to the problem:
- Product: Power user, casual user, churned user, new user
- Medical: Doctor, patient, caregiver, insurance
- Legal: Lawyer, affected party, regulator
- Technical: Security, performance, maintainability
- Content: Target audience, critic, competitor
Check personas.md for common perspective templates.
Synthesis
After collecting perspectives:
- Identify conflicts — Where do perspectives disagree?
- Find common ground — What do all perspectives agree on?
- Weigh by stakes — Safety concerns > preferences
- Decide — Make a call, document trade-offs
Check synthesis.md for conflict resolution patterns.
Learning User Preferences
Track when divergence helped vs. was overkill. Adapt:
- User who values speed → diverge less, ask before spawning
- User who values thoroughness → diverge more proactively
Related: For routing tasks to sub-agents, see delegate. For iterating until success, see loop.
1---2name: diverge3description: Spawn multiple perspectives to evaluate a problem before converging on a solution.4---5
6## When to Diverge
7
8Diverge when a single viewpoint isn't enough:
9- **Product decisions** — Different user types have conflicting needs
10- **Safety-critical** — Need doctor, lawyer, security expert perspectives
11- **Creative work** — Multiple aesthetic directions before choosing
12- **Complex trade-offs** — No obvious "right" answer
13
14Don't diverge for:
15- Simple tasks with clear answers
16- Speed-critical requests
17- Tasks where you already have high confidence
18
19## How Many Perspectives
20
21| Complexity | Perspectives | Examples |
22|------------|--------------|----------|
23| Low | 2-3 | Quick sanity check |
24| Medium | 3-5 | Product feature review |
25| High | 5-7 | Safety-critical, legal |
26
27**Rule:** Enough to cover blind spots, not so many you can't synthesize.
28
29## Choosing Perspectives
30
31Match perspectives to the problem:
32
33- **Product:** Power user, casual user, churned user, new user
34- **Medical:** Doctor, patient, caregiver, insurance
35- **Legal:** Lawyer, affected party, regulator
36- **Technical:** Security, performance, maintainability
37- **Content:** Target audience, critic, competitor
38
39Check `personas.md` for common perspective templates.
40
41## Synthesis
42
43After collecting perspectives:
441. **Identify conflicts** — Where do perspectives disagree?
452. **Find common ground** — What do all perspectives agree on?
463. **Weigh by stakes** — Safety concerns > preferences
474. **Decide** — Make a call, document trade-offs
48
49Check `synthesis.md` for conflict resolution patterns.
50
51## Learning User Preferences
52
53Track when divergence helped vs. was overkill. Adapt:
54- User who values speed → diverge less, ask before spawning
55- User who values thoroughness → diverge more proactively
56
57---
58
59**Related:** For routing tasks to sub-agents, see `delegate`. For iterating until success, see `loop`.
johnalbertini14-glitch