Jeden Skill in Manus ausführen
mit einem Klick

Jeden Skill in Manus mit einem Klick ausführen

Loslegen

qa-and-fix

Sterne2

Forks0

Aktualisiert27. April 2026 um 22:18

Standalone QA investigation and fix workflow

Installation

Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.

In Manus ausführen

Quelle

norlanp

norlanp/dotfiles

GitHub-Repository öffnen Creator-Repositorys ansehen

Download

In Manus ausführen

Verwandte BerufeSOC

Basierend auf der SOC-Berufsklassifikation

Softwarequalitätssicherungsanalysten und -testerInformatik- und Mathematikberufe·SOC 15-1253

SKILL.md

readonly

name	qa-and-fix
description	Standalone QA investigation and fix workflow

QA and Fix

/qa-and-fix - Standalone QA + fixes

Model Tiers

T1 = deep review rigor, T2 = balanced execution/fixes, T3 = quick triage only.
Task tool mapping: subagent_type="general"; if effort variants exist use T1=high, T2=medium, T3=low, else enforce via prompt depth/evidence.

Agents

Orchestrator: Coordinate, Playwright fallback
QA Agent: Distinguished QA Engineer (25+ yrs) - investigate, test, validate
Fix Agent: Distinguished Engineer (25+ yrs) - root cause, fix

Flow

Phase 1: Setup

Ask: "What should I QA first?" with choices:
- feature flow (recommended)
- specific file/module
- recent changes
- full project sweep
Read README, arch docs

Phase 2: QA (delegated)

[T2:balanced] Spawn QA agent:

Persona: Distinguished QA Engineer (25+ yrs)
Scope: {user-defined}

Check: functional, code quality, integration, tests, docs, security/perf
Playwright (web): try MCP → fail? return NEEDS_PLAYWRIGHT + scenarios

Output: issues (CRITICAL/MEDIUM/LOW) w/ file:line, passed checks

If NEEDS_PLAYWRIGHT → orchestrator executes, reports back
Present report → "Proceed to fix now? (recommended: yes) [Y/n]"

Phase 3: Fix (delegated)

[T2:balanced] Spawn fix agent:

Persona: Distinguished Engineer (25+ yrs)
Issues: {from QA}

For each (CRITICAL→MEDIUM→LOW):
 - Root cause → design fix → implement → verify
 - Follow `AGENTS.md`/`agents.md` philosophy

Output: fixed, files modified, deferred, PLAYWRIGHT_VALIDATION if needed

If Playwright validation → execute directly
Loop max 3x if validation fails

Phase 4: Review

[T1:deep] Spawn two subagent_type="general" review tasks with personas: Distinguished Code Reviewer + Distinguished Architect → if CHANGES_NEEDED → fix agent → repeat

Phase 5: Output

Save to docs/hotfixes/{hotfix-id}/: qa-report.md, fix-report.md, review-summary.md

Responsibilities

Task	Owner
User interaction	Primary session
QA investigation	QA Agent
Fix implementation	Fix Agent
Playwright testing	QA Agent (primary fallback)
Code review	Review Agents

Mehr aus diesem Repository

gleiches Repository

brainstorm

norlanp/dotfiles

Interview the user relentlessly about a plan or design. Use when the user wants to stress-test a plan before building, or uses any 'grill' trigger phrases.

2026-06-222

improve-code

norlanp/dotfiles

Surface deepening opportunities in a codebase — shallow modules, seam leaks, missing locality — and grill through whichever one you pick. Architecture-pattern aware (monolith / distributed / serverless).

2026-06-222

docs

norlanp/dotfiles

Reconcile and validate all documents in the project's docs/ directory

2026-06-092

audit

norlanp/dotfiles

Full-project code audit workflow: audit then fix

2026-05-162

debug

norlanp/dotfiles

Systematic root cause analysis. No fixes without investigation

2026-05-042

engineer

norlanp/dotfiles

Principal Engineer persona. Production-first technical judgment

2026-04-272

name	qa-and-fix
description	Standalone QA investigation and fix workflow

QA and Fix

/qa-and-fix - Standalone QA + fixes

Model Tiers

T1 = deep review rigor, T2 = balanced execution/fixes, T3 = quick triage only.
Task tool mapping: subagent_type="general"; if effort variants exist use T1=high, T2=medium, T3=low, else enforce via prompt depth/evidence.

Agents

Orchestrator: Coordinate, Playwright fallback
QA Agent: Distinguished QA Engineer (25+ yrs) - investigate, test, validate
Fix Agent: Distinguished Engineer (25+ yrs) - root cause, fix

Flow

Phase 1: Setup

Ask: "What should I QA first?" with choices:
- feature flow (recommended)
- specific file/module
- recent changes
- full project sweep
Read README, arch docs

Phase 2: QA (delegated)

[T2:balanced] Spawn QA agent:

Persona: Distinguished QA Engineer (25+ yrs)
Scope: {user-defined}

Check: functional, code quality, integration, tests, docs, security/perf
Playwright (web): try MCP → fail? return NEEDS_PLAYWRIGHT + scenarios

Output: issues (CRITICAL/MEDIUM/LOW) w/ file:line, passed checks

If NEEDS_PLAYWRIGHT → orchestrator executes, reports back
Present report → "Proceed to fix now? (recommended: yes) [Y/n]"

Phase 3: Fix (delegated)

[T2:balanced] Spawn fix agent:

Persona: Distinguished Engineer (25+ yrs)
Issues: {from QA}

For each (CRITICAL→MEDIUM→LOW):
 - Root cause → design fix → implement → verify
 - Follow `AGENTS.md`/`agents.md` philosophy

Output: fixed, files modified, deferred, PLAYWRIGHT_VALIDATION if needed

If Playwright validation → execute directly
Loop max 3x if validation fails

Phase 4: Review

[T1:deep] Spawn two subagent_type="general" review tasks with personas: Distinguished Code Reviewer + Distinguished Architect → if CHANGES_NEEDED → fix agent → repeat

Phase 5: Output

Save to docs/hotfixes/{hotfix-id}/: qa-report.md, fix-report.md, review-summary.md

Responsibilities

Task	Owner
User interaction	Primary session
QA investigation	QA Agent
Fix implementation	Fix Agent
Playwright testing	QA Agent (primary fallback)
Code review	Review Agents