Hire an AI code performance investigator

Everything it can hand back

Your task sets the actual outputs. The specialist picks the methods that cover them.

• The slow moment written in one line, with the stages ranked by what they cost
• A list of ideas the numbers rule out, so you skip them
• The biggest bottleneck fixed in code, measured before and after
• A pull request that shows the timings

How it does the job well

• Read the existing code, its conventions and its tests before changing anything. Match the project's patterns.
• Make the smallest change that fully solves the request. Keep unrelated refactors out.
• Reproduce a bug before fixing it, and add a test that fails without the fix.
• Treat input from users, files and networks as untrusted. Validate it at the boundary.

Mistakes it avoids

• Claiming code works without running it or its tests.
• Inventing APIs, package versions or settings instead of checking the installed ones.
• Changing behaviour outside the request, such as reformatting whole files or renaming public interfaces.

What a good result looks like

An example, not your task: A page, report, job or query feels slow and the team is guessing why.

• Slow parts ranking. The slow moment in one line, a timing plan, stages ranked by cost and ideas the numbers rule out.
• Fix and measure. The before number, the cause found, the code change and the after number.
• Pull request. A short summary of the change, the timings before and after, and how risky it is to merge.
• HAND-BACK SUMMARY. Each step's file, any step not run and why, and every MISSING line in one list.

How the work is checked

• Each step passed its own checks, or the run stopped at the step that failed and says why.
• The steps ran in order, and each one used the result of the one before.
• Nothing was invented to fill a gap. Every gap is a MISSING: line.
• Nothing was sent, published or paid for without a clear yes.
• States which checks passed, failed or could not run, and names missing inputs.

These checks have not been run against a model yet. Your AI uses them to check its own work.

A second opinion

Before it hands back, your AI reviews the draft the way John Carmack would. This is a way of reviewing the work. It does not mean they wrote or endorse it.

Capabilities

• Rank the slow parts by cost. Writes the slow moment in one line, plans real timings, looks at the worst runs and ranks the fixes by the time they save.
• Run the timings (your turn). You run the timings it planned and note the numbers for the top stage.
• Fix the bottleneck and measure again. Measures the top stage in the code, finds the cause with a profiler or query plan, fixes the biggest cost and measures again.
• Write the pull request. Describes the change in a small picture, with the timings before and after and how hard it is to undo.

How the role works

1. Checks your task, context and available tools.
2. Picks the methods needed and prepares a work plan.
3. Carries out supported tasks, with each method's limits and review gates.
4. Checks the real deliverables and corrects supported faults.
5. Hands back files, evidence and unresolved gaps.

Your specialist's working setup

Download the base specialist without a task form or account. README.md and SETUP.md explain how to use the folder. SOUL.md defines its scope. INTAKE.md asks missing questions inside your AI. PLAYBOOK.md, SKILLS.md, BEST_PRACTICES.md, CHECKS.md and HANDOFF.md guide the work and its checked deliverables.

Complete associated skills, supporting files and licences are included. Choose only the methods needed for the actual task. Keep context, working files and accepted feedback in your AI project.

Use this specialist through MCP or the API

Call prepare_specialist with role_id find-what-is-slow-fix-it-and-show-the-numbers only. No task is required to load it. Your AI asks what job you need done, what you want back and which relevant context it can use.

For task-led hiring, call find_specialist, explain the suggested specialists and let the user accept or change the team.

Complete ZIP: Download base specialist. API: GET /api/v1/specialists/find-what-is-slow-fix-it-and-show-the-numbers/pack. The host AI supplies its own model, tools and execution; a pack does not connect accounts or run a task.

Inputs and helpful context

The specialist starts from the task brief. Only active methods require their inputs.

For the primary method, when used

• What is slow, when it happens and about how long it takes
• Access to your project code
• Logs, timings or a slow-query list (optional)
• Your target speed (optional)

Tools and access

• Any good chat AI: Ranking costs and planning measurements is text work any strong chat AI does.
• A profiler or a stopwatch: A real timing tool or clock gives the numbers every claim in the report depends on.

Functions and deliverables

• Code performance investigator: Find the slow code path and measure it. Make a focused fix and write a pull request with before-and-after timings.
  • The slow moment written in one line, with the stages ranked by what they cost
  • A list of ideas the numbers rule out, so you skip them
  • The biggest bottleneck fixed in code, measured before and after
  • A pull request that shows the timings
• Slowdown investigation planner: Describe what feels slow. Get a measurement plan, worst-case review and ranked fixes where timings support them.
  • The slow moment in one sentence
  • A plan for what to time, and with what
  • A worst-case view next to the average
  • A ranked list of what to fix, biggest cost first, and a list of dropped ideas
• Slow code fixer: Share slow code and a repeatable workload. Get a measured bottleneck, proposed fixes and before-and-after checks.
  • The bottleneck, named
  • A fix aimed at the slowest part
  • Before and after timings
  • A short note on what changed and why
  Source: sickn33.
• Pull request writer: Turn your code changes into a pull request description with a picture, proof it works, and a merge risk note that reviewers can scan fast.
  • Pull request body in a set template
  • A diagram, tree or code sketch of the change
  • Before and after evidence section
  • Merge risk rating
  Source: Matt Pocock.
• AI project context organiser: Share your project and next task. Get a short rules file, a focused file list and a clear handoff for later sessions.
  • A ready rules file for your project
  • A list of which files to load for each task
  • Rules for when to reset or trim the session
  • A checklist for spotting drifting answers
  Source: Addy Osmani.
• John Carmack: Find a bug, slow step or excess scope through John Carmack’s recorded approach. Follow the actual run, measure first and expose hidden state.
  • Your problem worked on the way John Carmack would: their questions, their method, their push-back
  • The answer or the work, in John Carmack's voice, using their ideas by name
  • Anything that goes beyond their public record, marked
  • One suggested next step

Tasks you can use it for

• A page, report, job or query feels slow and the team is guessing why.
• A speed fix was tried and nobody can say if it worked.
• You need numbers to show a reviewer or a client.

Limits to check before starting

• The slowness is a one-off cost, such as a first load: run a profiler on a single run instead.
• You need a general code review, not a speed fix: ask a developer on your team to review the code.
• You have no access to the code: send your timings and notes to the developer who owns it.

Best way to use this role

Give it one clear outcome, the actual files and a definition of done. Keep the work in one AI project so it can reuse checked context and earlier files.

• Download the handoff pack. It holds HANDOFF.md plus one skill file for each step.
• Open your project in Claude Code, give it the folder and say "Follow HANDOFF.md".
• Describe what is slow and how long it takes. Add any logs you have.
• Run the timings the first step plans. Tell it the numbers for the top stage.
• Check the before and after figures in the pull request on a run of your own.

Read it as data

Read the complete role setup as JSON. Connect your AI to the AI Blocks MCP library and read aiblocks://roles/find-what-is-slow-fix-it-and-show-the-numbers for its role instructions and pinned sources.

This is a prepared AI role. No task has run. Your AI carries out authorized work using its available tools and reports the actual deliverables and checks.