Plan Review Architecture

Plan Review Architecture

Review architecture after scope is agreed.

Guardrails

• Evaluate only architecture affected by the agreed scope.
• Prefer existing boundaries and local patterns unless they cannot support the goal.
• Recommend new abstractions only for current complexity, not hypothetical future cases.
• Tie every architecture concern to a failure mode, integration risk, or verification step.

Evaluate

• Component boundaries and ownership.
• Dependency graph and coupling.
• Data flow, state transitions, and side effects.
• API and integration contracts.
• Security and privacy boundaries.
• Rollout, rollback, migrations, and partial deployment states.
• Observability: logs, metrics, traces, dashboards, and alerting.
• One realistic production failure scenario for each new codepath or integration.

Diagrams

Require ASCII diagrams for non-trivial:

• Data flows
• State machines
• Processing pipelines
• Dependency graphs
• Decision trees

Recommend inline diagram comments only where they will remain close to complex logic.

Output

Return architecture findings ordered by severity. For each issue include:

• Finding
• Evidence with file references when available
• Recommendation
• Options if user judgment is needed
• Failure mode if ignored

---

> Install: `npx skills add ChristopherAlphonse/calphonse-skills --skill plan-review-architecture`

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: ChristopherAlphonse
• Source: ChristopherAlphonse/calphonse-skills
• License: MIT
• Homepage: https://www.skills.sh/christopheralphonse/calphonse-skills

Install and usage instructions live in the source repository linked above.