Skip to content
rd-skills
All skills

profiling

FoundationJIT only

task-agent/review-agent: use when CPU, memory, I/O, database, network, rendering, or cost needs measured bottleneck evidence; skip without a profiling need.

Group
Reliability & operations
Supported roles
MainAnalysisTaskReview
Source
SKILL.md

Use when

Trigger signals that make this skill the right owner

  • profile CPU memory IO database network rendering cloud cost unit economics and bottlenecks

Do not use when

Anti-triggers — as binding as the triggers above

  • no task-local profiling decision is required

Required inputs

What must be supplied before this skill can decide anything

  • current task contract
  • selected primary Professional Skill
  • task-local trigger evidence

Output contract

What this skill owes the next role

  • profiling evidence with symptom and hypothesis, representative comparable workload, measured bottleneck and re-profile, resource/cost transfer, overhead boundary, correctness guard, and owned sensitive-artifact lifecycle

Escalation signals

When this skill must hand the decision back

  • Guessing at bottlenecks or cost drivers wastes effort and can worsen real constraints

Loaded by

This skill never owns a task. These professionals may pull it in as Layer 3.

Targeted references

Each reference carries its own load condition. None of them enter context by default.

ReferenceLoad whenDo not load whenRequired by
benchmarks-and-patterns.mdbenchmark-patternCompeting profiler signal workload or bottleneck patterns remain viable.One authorized representative method directly tests the hypothesis and root rules resolve the decision.TaskReview
checklist.mddecision-checklistSeveral comparison overhead transfer correctness or artifact-lifecycle decisions must close together.One bounded measurement has comparable evidence an owner a stop path and no unresolved artifact lifecycle.TaskReview
evidence-patterns.mdevidence-patternHypothesis representative comparison bottleneck re-profile transfer overhead correctness or sensitive-artifact freshness claims need proof.Representative matched final measurements correctness checks and owned artifact cleanup close the claims.TaskReview