# Performance

> Review performance and observability — bottlenecks, baselines, metrics, tracing. Use only when the user explicitly asks for a performance or observability review or runs /performance.

- Skill: `limonequantistico/performance` (Agent Skill)
- Install (CLI): `npx skillmds@latest add limonequantistico/performance`
- Raw SKILL.md: https://api.skillmd.com/api/skills/limonequantistico/performance/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: limonequantistico (https://skillmd.com/u/limonequantistico)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/limonequantistico/performance

---


You are reviewing **performance and observability**: how the app behaves under load, what's being measured, and whether operators can see production issues coming.

**Scope note:** application logging — what gets recorded, at what level, with what context — belongs to `/logs`. Flag a logging gap if you trip over one, then point at `/logs` rather than designing the contract here.

**Custom instructions:** If the user included extra instructions when invoking this skill, treat them as overriding or extending the guidance below.

**Read (as applicable):**

- `.docs/seed.md` — latency-sensitive flows, scale expectations, SLAs if any
- `.docs/tech-stack.md` — framework, hosting, suggested monitoring tools
- Code: hot paths (DB queries, N+1 patterns, large payloads), caching, background jobs, API routes, client bundles
- Existing metrics, tracing, and error reporting (Sentry, OpenTelemetry, APM, etc.)

**Your task:**

1. **Performance** — Identify likely bottlenecks in the **current** implementation (not hypothetical). Prefer concrete references (file + pattern): missing indexes, unbounded queries, sync work on request thread, heavy client JS, missing pagination, etc.

2. **Baselines** — Suggest what to measure (p95 latency, error rate, job queue depth) and where — only what fits this stack.

3. **Tracing & metrics** — Gaps: no structured errors, no health check, no uptime probe, no dashboards, no trace context across service boundaries. Recommend minimal additions aligned with the tech stack.

4. **Prioritized actions** — Short list: highest impact first, each item actionable in one sitting.

**RULES:**

- Do not recommend a new vendor or library unless the user asks or the stack already allows it; otherwise flag once for user decision.
- Focus on shipped code paths, not future features.

**Logging:** On success, append to `.docs/changelog.md` per `.docs/changelog-spec.md`: `- YYYY-MM-DD HH:mm ran /performance — [brief summary]`.