Skill Router
Given a problem, map it to the right skills and sequence them for maximum leverage.
How to Use
Describe your problem in plain language. The router will:
- Classify the domain (delivery, architecture, quality, people, metrics)
- Return a ranked list of applicable skills
- Suggest a sequencing order with rationale
- Highlight skill dependencies or synergies
Domain → Skill Map
Delivery & Flow
| Trigger |
Skills |
| Code is slow to ship |
continuous-delivery → feature-flags → flow-metrics-analysis |
| PRs take too long |
code-review → pair-programming → flow-metrics-analysis |
| Deployment is risky |
continuous-delivery → feature-flags → observability |
| Work piles up mid-sprint |
sprint-health-check → flow-metrics-analysis → backlog-refinement |
| End of sprint reflection |
sprint-retrospective → flow-metrics-analysis |
Architecture & Design
| Trigger |
Skills |
| Migrating off a monolith |
strangler-fig → monolith-to-services → event-driven-design → architecture-decision-record |
| System is hard to change |
refactoring → technical-debt-prioritization → architecture-decision-record |
| Building a new API |
api-design → documentation → security-review |
| Shipping an AI agent |
agent-quality-patterns → ci-cd-pipeline-analysis → observability |
| Database changes are scary |
database-migration → observability → feature-flags |
| Need to record a key decision |
architecture-decision-record |
Quality & Safety
| Trigger |
Skills |
| No tests / hard to test |
test-driven-development → refactoring → continuous-delivery |
| Security concerns on a feature |
security-review → code-review → observability |
| Production incident |
incident-retrospective → observability → on-call-handoff |
| Hard to debug |
debugging → observability |
QA / Quality Engineering
| Trigger |
Skills |
| Writing a test strategy for a feature or release |
test-strategy → acceptance-criteria → test-driven-development |
| Bug found — need a high-quality report |
bug-reporting → acceptance-criteria |
| Exploring edge cases / finding what scripted tests miss |
exploratory-testing → bug-reporting |
| Deciding what acceptance criteria to put on a story |
acceptance-criteria → backlog-refinement |
| Breaking down a large epic into sprint-ready stories |
story-slicing → backlog-refinement → acceptance-criteria |
| Setting up or reviewing the test pipeline |
ci-cd-pipeline-analysis → test-strategy |
| AI/LLM test automation quality |
agent-quality-patterns → test-strategy |
| Evaluating or adopting an MCP server |
mcp-safety-review → agent-quality-patterns |
| Agent gives generic, unhelpful answers |
context-engineering → agent-quality-patterns |
| Need to write CLAUDE.md / AGENTS.md / copilot-instructions.md |
context-engineering |
| UI/UX accessibility audit or WCAG compliance |
accessibility-testing |
People & Teams
| Trigger |
Skills |
| Team struggling to deliver |
team-topology → flow-metrics-analysis → sprint-health-check |
| Planning next quarter |
roadmap-prioritization → backlog-refinement → technical-debt-prioritization |
| Growing the team |
hiring-leveling → team-topology |
| New engineer joining the team |
engineer-onboarding → one-on-one-coaching |
| Preparing or running a 1:1 |
one-on-one-coaching |
| Writing or reviewing OKRs |
okr-engineering-alignment → engineering-metrics |
| Evaluating a new tool or vendor |
vendor-evaluation |
| Defining or improving tech direction |
technical-strategy → roadmap-prioritization |
| On-call handoff |
on-call-handoff → incident-retrospective |
| Tech debt is blocking features |
technical-debt-prioritization → refactoring → engineering-metrics |
Platform & Reliability
| Trigger |
Skills |
| Setting up or improving an internal developer platform |
platform-engineering → developer-experience |
| Developers complaining about slow CI / friction |
developer-experience → ci-cd-pipeline-analysis |
| Defining reliability targets or error budgets |
slo-sla-management → observability |
| Service is slow or unreliable under load |
performance-optimization → observability → slo-sla-management |
| Testing service resilience |
chaos-engineering → observability |
| Microservices integration testing |
contract-testing → test-strategy |
| Slow database queries |
database-optimization |
Routing Template
When asked "where do I start?", respond with:
Problem: [restate the problem]
Domain: [delivery / architecture / quality / people / metrics]
Recommended skill sequence:
1. [skill-name] — [why first]
2. [skill-name] — [depends on output from step 1]
3. [skill-name] — [optional, if complexity warrants]
Key synergy: [note any skills that reinforce each other]
Skip if: [condition under which a skill can be omitted]
Skill Synergy Notes
strangler-fig + monolith-to-services: Always pair these. Strangler-fig is the migration pattern; monolith-to-services provides the decomposition strategy.flow-metrics-analysis + engineering-metrics: Flow metrics are the what (cycle time, throughput); engineering metrics are the why (DORA, team health). Use both for quarterly reviews.refactoring + test-driven-development: Refactoring without tests is risky. TDD first, then refactor safely.architecture-decision-record + any arch skill: Any major decision from api-design, event-driven-design, or monolith-to-services should produce an ADR.observability + continuous-delivery: You can't safely ship fast without knowing what's happening in production.security-review + code-review: Security review surfaces threats; code review enforces them. Run in the same PR flow.
References