Skip to content
rd-skills
All skills

architecture-impact-reviewer

ProfessionalRuntime top level

Use analysis-agent for module-boundary or dependency-direction analysis, or review-agent for independent assessment of a bounded architecture artifact. Skip isolated owner-internal edits with no structural impact.

Group
Impact, domain and architecture
Routing mode
evidence-only
Supported roles
MainAnalysisTaskReview
Source
SKILL.md

Use when

Trigger signals that make this skill the right owner

  • module boundary change
  • dependency direction risk
  • owner-internal implementation structure reuse or deliberate separation remains unresolved

Do not use when

Anti-triggers — as binding as the triggers above

  • isolated owner internal edit after placement is fixed
  • no structural impact

Required inputs

What must be supplied before this skill can decide anything

  • desired behavior
  • module and dependency boundary

Output contract

What this skill owes the next role

  • placement decision
  • architecture verdict
  • structural risk

Escalation signals

When this skill must hand the decision back

  • scope, authority, or material risk exceeds the selected task contract

Targeted references

Each reference carries its own load condition. None of them enter context by default.

ReferenceLoad whenDo not load whenRequired by
placement-and-ownership.mdtargetedStructure responsibility ownership reuse or placement remains openOwner public/private surface and smallest viable placement are fixed by current evidenceAnalysisReview
consumer-and-data-impact.mdtargetedPublic or indirect consumer or authoritative-data impact remains openNo consumer contract compatibility versioning migration or data-owner boundary changesAnalysisReview
dependency-topology-and-enforcement.mdtargetedDependency direction topology or durable enforcement remains openCurrent source edges topology owner and enforcement evidence already fix the boundaryAnalysisReview
reversibility-evolution-and-proof-limits.mdtargetedReversal coexistence migration evolution or architecture proof limits remain openCurrent evidence fixes reversibility evolution residual risk and next ownerAnalysisReview
checklist.mddecision-checklistA bounded L2 mode needs compact checks for its triggered placement, ownership, dependency, consumer, data, topology, or reversibility riskThe root contract is enough or targeted proof fields are requiredAnalysisReview
index.mdindexcompeting architecture impact reviewer references require dependency, conflict, or output-fragment selectionthe architecture impact reviewer root or a task-named reference already resolves selectionAnalysisReview
solution-optimality.mdtargetedAn architecture boundary, ownership, dependency, topology, or operational-responsibility decision has a material alternativeThe change is owner-internal with no structural impact or current evidence already fixes placementAnalysisReview