Performance hotspot check
Skill vishnujchandran/.agents/skills/performance-hotspot-check
A curated collection of reusable AI agent skills, playbooks, and prompts for security audits, engineering workflows, and automation.
npx -y skills add vishnujchandran/.agents --skill performance-hotspot-checkAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Evidence-first performance analysis skill that identifies high-impact bottlenecks in app code, database access, and API calls with safe, minimal-diff optimization plans.
SKILL.md
2.1 KB, 425 tokens by cl100k_base, as published. Nobody here has run it
performance-hotspot-check (v2.0)
Find real bottlenecks and optimize highest-impact hotspots first.
Trigger Signals
Use when user asks to:
- find slow app/API paths
- diagnose latency spikes
- optimize DB/API-heavy workflows
- reduce response time, compute cost, or build/runtime overhead
Modes
analyze-only(default): bottleneck report + optimization planpatch-safe: apply low-risk optimizations after user confirmation
Baseline First (Required)
Capture current metrics before recommending fixes:
- latency (p50/p95/p99 where possible)
- DB query count/time for hot requests
- external API call count/time
- CPU/memory symptoms if available
What It Checks
- Repeated expensive computation on hot paths
- N+1 query patterns and missing indexes
- Unbounded pagination/filtering/sorting
- Serial external calls that can be reduced/batched/cached
- Over-fetching and heavy payload construction
- Blocking sync operations in latency-sensitive paths
Workflow
- Identify critical user/API paths and frequency.
- Collect baseline metrics and evidence.
- Locate bottlenecks in code, DB, and external calls.
- Rank by impact (
latency × frequency × cost). - Propose minimal, low-risk optimizations.
- Define before/after verification plan.
Output Contract
Return exactly:
- Hotspot summary with baseline evidence
- Findings table:
priority | area | file | bottleneck | impact | fix - Top 5 optimizations first
- Measurement plan (before/after metrics)
- Rollback notes for risky changes (if any)
Priority
- P0: major user-facing latency/cost bottlenecks
- P1: medium-impact inefficiencies on common paths
- P2: low-impact cleanup opportunities
Guardrails
- No optimization without evidence
- Never trade correctness/security for speed
- Prefer minimal diffs and reversible changes
- Call out assumptions explicitly when profiling data is incomplete