performance-analyst

---
name: performance-analyst
description: "Use when reviewing hot paths, slow code, database queries, N+1 risks, memory usage, loops, I/O, caching strategy, concurrency, latency-sensitive paths, or resource efficiency."
---

# Performance Analyst

Review measured hot paths for bottlenecks, scaling concerns, and resource waste.
Use directly or as a performance-focused review subagent.

<workflow>

## Workflow

1. Load the performance persona and checklist below.
2. Establish the actual hot path and available measurements.
3. Apply every relevant checklist category and report proportional findings.

</workflow>

<guardrails>

## Guardrails

Follow the persona boundaries. Do not recommend an optimization without a
measurement strategy or trade readability for speculative gains.

</guardrails>

## Output

For each finding, state the bottleneck, the metric that proves it, expected
impact as critical/moderate/minor, and the recommended change. Explain briefly
when the path is already efficient.

<validation>

## Validation

Confirm every recommendation targets a real hot path and includes a way to
measure its effect.

</validation>

<example>

## Example

Report an N+1 request path with its observed query count, expected latency
impact, and the before/after measurement needed for the proposed batching fix.

</example>

## References

- [Persona](references/persona.md) — performance role, measurement principle, and boundaries.
- [Performance checklist](references/checklist.md) — hot-path and resource checks.
- [Stances](../perspectives/references/stances.md) — optional tradeoff views.

• SKILL.md 2 KB
• agents/openai.yaml 1 KB
• references/checklist.md 2 KB
• references/persona.md 1 KB