Using Syarif Laravel Standards

Apply silently.

1. Smallest safe change
2. Infer risk: LOW / MEDIUM / HIGH.
3. Preserve unrelated behavior.
4. Select relevant guidance.
5. Verify proportionally.
6. Stop when complete.

Memory preflight is conditional infrastructure, not specialist activation. Before broad exploration, decide whether prior project/session/workflow/decision context could materially affect correctness. If yes, prefer the active syarif-memory-management MCP memory_auto; otherwise run node skills/memory-management/scripts/memory.mjs auto --cwd <project-root> --query "<task intent>" --limit 5 when Node/file access exists; if unavailable, continue without memory. Skip memory for self-contained syntax, isolated-helper, or generic documentation tasks. Retrieved memory is sparse orientation only: verify against current code/config/docs, and current code/config wins.

Smallest safe change: choose the minimal change satisfying required behavior, correctness, contracts, and safety. Do not omit required branches, conditions, validation, authorization, lifecycle, error handling, or explanation to reduce output length. Be concise but complete; include enough context for semantic correctness. Prefer a coherent path; combine elements when correctness requires them.

LOW: local change. No new abstraction. Minimal verification. No memory unless needed. MEDIUM: trace affected path. Preserve contracts. Root-cause where non-obvious. Targeted regression check. HIGH: inspect security/data/concurrency/auth/migration/financial risks. Failure paths. Affected regression surface. Explicit remaining uncertainty when meaningful.