| name | Write Engineering Strategy |
| description | Help craft effective engineering strategies using the Diagnosis, Policies, Actions framework. Use when writing a strategy document, starting a new leadership role, aligning a team on direction, or preparing for planning season. |
The Framework
1. Diagnosis
What's actually happening?
A good diagnosis:
- Names the core challenge specifically
- Is grounded in evidence
- Acknowledges constraints
- Explains why the problem exists
Questions to ask:
- What's the symptom you're seeing?
- What data supports this observation?
- How long has this been happening?
- What's caused previous attempts to fail?
- What constraints shape the situation?
Bad: "We need to move faster"
Good: "Feature delivery has slowed because tech debt in the payments system requires every change to navigate complex legacy code, while our testing infrastructure can't keep up with deploy frequency."
2. Guiding Policies
How will you approach this?
Good policies:
- Flow logically from the diagnosis
- Provide clear direction
- Help teams make autonomous decisions
- Are memorable
Questions to ask:
- Given the diagnosis, what approach makes sense?
- What principles should guide decisions?
- What tradeoffs are you making?
- How will teams know if they're on track?
Bad: "We will prioritize quality"
Good: "We will invest 30% of capacity in platform work until the payments system can be modified safely in under a week."
3. Coherent Actions
What specifically will you do?
Good actions:
- Implement the policies
- Are specific and executable
- Are coordinated across teams
- Are sequenced logically
Questions to ask:
- What concrete initiatives support this policy?
- Who owns each initiative?
- What's the sequence and dependencies?
- How will you measure progress?
Bad: "Reduce tech debt"
Good: "Q1: Extract payments to separate service. Q2: Add integration tests for 80% of flows. Q3: Migrate traffic to new service."
Process
- Gather context: Review current state, talk to stakeholders
- Draft diagnosis: What's the core challenge?
- Test diagnosis: Does it ring true? What's missing?
- Develop policies: What approach addresses the diagnosis?
- Define actions: What implements the policies?
- Identify tradeoffs: What are you not doing?
- Set metrics: How will you know it's working?
- Review and iterate: Get feedback, refine
Common Pitfalls
- All vision, no diagnosis
- All actions, no policies
- Too many priorities
- Not actually deciding
- Writing it and forgetting it
Templates
Reference: templates/engineering-strategy.md
Output
Help them produce:
- Clear diagnosis of the core challenge
- 2-4 guiding policies
- Specific actions with owners and sequencing
- Explicit tradeoffs (what you're not doing)
- Measurable success criteria