t1-performance-evidence

Étoiles0

Forks0

Mis à jour14 juillet 2026 à 01:00

T1 measurement-driven performance optimization playbook. Use when a job claims latency, throughput, rendering, query, memory, or resource concerns.

Installation

Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.

Source

lebobo88

lebobo88/orchestrator

Ouvrir le dépôt GitHub Voir les dépôts du créateur

Téléchargement

Métiers associésSOC

Basé sur la classification professionnelle SOC

Développeurs de logicielsProfessions informatiques et mathématiques·SOC 15-1252

SKILL.md

readonly

name	t1-performance-evidence
description	T1 measurement-driven performance optimization playbook. Use when a job claims latency, throughput, rendering, query, memory, or resource concerns.
user-invocable	false

Performance evidence

Define the user-visible or operational metric, workload, environment, and acceptable regression limit.
Capture a reproducible baseline before changing production code. Trace the likely bottleneck and distinguish CPU, memory, I/O, network, database, rendering, and algorithmic causes.
Form a bounded hypothesis; make one small change at a time under the normal TDD loop.
Measure before/after with the same method. Do not claim improvement without comparable evidence.
Check correctness, error behavior, cache invalidation, resource lifecycle, and worst-case/edge inputs. Report noise, limits, and any tradeoff honestly.

Plus depuis ce dépôt

même dépôt

browser-validator-core

lebobo88/orchestrator

Screenshot-first visual end-to-end validation policy for browser-rendered UI and webview jobs after verifier approval.

2026-07-180

build

lebobo88/orchestrator

Default engineering workflow. Classify a technical request, route it through the native T1/T2/T3 engineering fleet, and return only independently verified completion to the orchestrator.

2026-07-180

claude-operations

lebobo88/orchestrator

Safely operate Claude Code's opt-in advanced capabilities—sessions, worktrees, agent teams, Skills, MCP, plugins, hooks, Channels, schedules, goals, and Agent SDK integrations—when the user explicitly asks for one. Do not use for an ordinary build request.

2026-07-170

cross-vendor-judging

lebobo88/orchestrator

Routes selective local Codex plan, code, verification, and visual judgment with evidence-backed blocking and strict cost caps.

2026-07-170

design-core

lebobo88/orchestrator

Shared evidence, accessibility, context, and handoff rules for the local design-agent fleet.

2026-07-170

design-direction

lebobo88/orchestrator

Produces contextual creative theses and bounded visual directions without generic model defaults.

2026-07-170

name	t1-performance-evidence
description	T1 measurement-driven performance optimization playbook. Use when a job claims latency, throughput, rendering, query, memory, or resource concerns.
user-invocable	false

Performance evidence

Define the user-visible or operational metric, workload, environment, and acceptable regression limit.
Capture a reproducible baseline before changing production code. Trace the likely bottleneck and distinguish CPU, memory, I/O, network, database, rendering, and algorithmic causes.
Form a bounded hypothesis; make one small change at a time under the normal TDD loop.
Measure before/after with the same method. Do not claim improvement without comparable evidence.
Check correctness, error behavior, cache invalidation, resource lifecycle, and worst-case/edge inputs. Report noise, limits, and any tradeoff honestly.