🎯

error-handling

🎯Skill

from srstomp/pokayokay

VibeIndex|
What it does

Validates, categorizes, and automatically routes software errors to appropriate resolution strategies based on predefined error taxonomies and project-specific handling protocols.

πŸ“¦

Part of

srstomp/pokayokay(24 items)

error-handling

Installation

npxRun with npx
npx pokayokay
Add MarketplaceAdd marketplace to Claude Code
/plugin marketplace add srstomp/pokayokay
Install PluginInstall plugin from marketplace
/plugin install pokayokay@srstomp-pokayokay
npxRun with npx
npx @stevestomp/ohno-cli init
npxRun with npx
npx pokayokay doctor
Claude Desktop ConfigurationAdd this to your claude_desktop_config.json
{ "mcpServers": { "ohno": { "command": "npx", "args": ["@steve...
πŸ“– Extracted from docs: srstomp/pokayokay
5Installs
-
AddedFeb 4, 2026

Skill Details

SKILL.md

Overview

```

__ __

____ ____ / /______ _ __ ______ / /______ ___ __

/ __ \/ __ \/ //_/ __ / / / / / __ \/ //_/ __ / / / /

/ /_/ / /_/ / ,< / /_/ / / /_/ / /_/ / ,< / /_/ / /_/ /

/ .___/\____/_/|_|\__,_/ \__, /\____/_/|_|\__,_/\__, /

/_/ /____/ /____/

```

# pokayokay

AI-assisted development orchestration - A Claude Code plugin that orchestrates AI-assisted development sessions with configurable human oversight, bridging the gap between hands-on control and full automation through skills, hooks, agents, and integration with ohno for task management.

Features

  • PRD to Tasks - Automatically break down requirements into epics, stories, and tasks
  • Orchestrated Sessions - Work across multiple sessions without losing context
  • Human Checkpoints - Choose your autonomy level: supervised, semi-auto, or autonomous
  • Multi-Dimensional Auditing - Verify accessibility, testing, security, docs, and observability
  • 23 Specialized Skills - Route work to domain experts automatically
  • Spike Protocol - Time-boxed investigations with mandatory decisions

Prerequisites

  • Claude Code v1.0.0 or later
  • Node.js v18 or later (for ohno CLI)
  • Git (for version control integration)

Installation

The easiest way to install is with the setup wizard:

```bash

npx pokayokay

```

This interactive wizard will:

  1. Install the pokayokay Claude Code plugin
  2. Configure the ohno MCP server
  3. Initialize ohno in your project
  4. Optionally set up kaizen integration

Run npx pokayokay doctor anytime to verify your installation.

Manual Installation

Click to expand manual steps

```bash

# 1. Add the marketplace (one-time setup)

claude plugin marketplace add srstomp/pokayokay

# 2. Install the plugin

claude plugin install pokayokay@srstomp-pokayokay

```

Or from inside Claude Code REPL:

```

/plugin marketplace add srstomp/pokayokay

/plugin install pokayokay@srstomp-pokayokay

```

#### Required: ohno MCP Server

Add to your MCP configuration (~/.claude/settings.json):

```json

{

"mcpServers": {

"ohno": {

"command": "npx",

"args": ["@stevestomp/ohno-mcp"]

}

}

}

```

#### Initialize ohno

```bash

npx @stevestomp/ohno-cli init

```

Quick Start

```bash

# 1. Run setup wizard (if not done already)

npx pokayokay

# 2. Restart Claude Code to activate MCP server

# 3. Plan from a PRD

/pokayokay:plan docs/prd.md

# 4. View kanban board

npx @stevestomp/ohno-cli serve

# 5. Start working

/pokayokay:work supervised

# 6. Audit completeness

/pokayokay:audit --full

```

Commands

Core Workflow

| Command | Description |

|---------|-------------|

| /pokayokay:plan | Analyze PRD and create tasks with skill routing |

| /pokayokay:revise [--direct] | Revise existing plan with impact analysis |

| /pokayokay:work [mode] | Start/continue work session (supervised/semi-auto/autonomous) |

| /pokayokay:audit [feature] | Audit feature completeness across 5 dimensions |

| /pokayokay:review | Analyze session patterns and skill effectiveness |

| /pokayokay:handoff | Prepare session handoff with context preservation |

| /pokayokay:hooks | View and manage hook configuration |

| /pokayokay:worktrees | List, cleanup, switch, or remove worktrees |

Ad-Hoc Work

| Command | Description |

|---------|-------------|

| /pokayokay:quick | Create task and immediately start working |

| /pokayokay:fix | Bug fix with diagnosis workflow |

| /pokayokay:spike | Time-boxed technical investigation |

| /pokayokay:hotfix | Production incident response |

Development

| Command | Description |

|---------|-------------|

| /pokayokay:api | API design - REST/GraphQL patterns |

| /pokayokay:arch | Architecture review and refactoring |

| /pokayokay:db | Database schema and migrations |

| /pokayokay:test | Testing strategy and implementation |

| /pokayokay:integrate | Third-party API integration |

| /pokayokay:sdk | SDK creation and extraction |

Infrastructure & Quality

| Command | Description |

|---------|-------------|

| /pokayokay:cicd | CI/CD pipeline creation and optimization |

| /pokayokay:security | Security audit and vulnerability scanning |

| /pokayokay:observe | Logging, metrics, and tracing |

Research & Documentation

| Command | Description |

|---------|-------------|

| /pokayokay:research | Extended technical research |

| /pokayokay:docs | Technical documentation |

Skills

The plugin includes 23 specialized skills that are automatically loaded based on task type:

Backend & API

  • api-design - REST/GraphQL endpoint design
  • api-integration - Third-party API consumption
  • database-design - Schema design, migrations, optimization
  • architecture-review - Code structure, module boundaries

DevOps & Infrastructure

  • ci-cd-expert - GitHub Actions, GitLab CI, deployment strategies
  • observability - Logging, metrics, tracing, alerting

Quality & Security

  • testing-strategy - Test architecture, coverage, E2E patterns
  • security-audit - OWASP Top 10, dependency scanning

Investigation

  • spike - Time-boxed technical investigation (2-4 hours)
  • deep-research - Multi-day technology evaluation

Planning

  • plan-revision - Guided plan revision with impact analysis

Audit Dimensions

The /pokayokay:audit command checks 5 dimensions:

| Dimension | Levels | Description |

|-----------|--------|-------------|

| Accessibility | L0-L5 | Is the feature user-accessible? |

| Testing | T0-T4 | Test coverage and types |

| Documentation | D0-D4 | Code comments to user docs |

| Security | S0-S4 | Input validation to hardened |

| Observability | O0-O4 | Logging to full telemetry |

```bash

/pokayokay:audit # Quick (accessibility only)

/pokayokay:audit --dimension testing # Specific dimension

/pokayokay:audit --full # All dimensions

```

Work Modes

| Mode | Task | Story | Epic |

|------|------|-------|------|

| supervised | PAUSE | PAUSE | PAUSE |

| semi-auto | log | PAUSE | PAUSE |

| autonomous | skip | log | PAUSE |

Parallel Execution

Run multiple tasks simultaneously for faster throughput:

```bash

# Run up to 3 tasks in parallel

/pokayokay:work semi-auto --parallel 3

# Short form

/pokayokay:work -p 3

```

How it works:

  • Coordinator dispatches N implementer agents in a single message
  • Each agent works independently with fresh context
  • Results processed as they complete
  • Dependency graph prevents unsafe parallelization

Recommended settings:

  • Default: 1 (sequential, safest)
  • Independent tasks: 2-3
  • Large backlog: 3-4
  • Maximum: 5

Tradeoffs:

  • Higher token usage (N concurrent contexts)
  • Potential git conflicts (auto-resolved when possible)
  • No shared learning between parallel agents

Worktree Isolation

Tasks automatically run in isolated git worktrees based on type:

| Task Type | Behavior | Override |

|-----------|----------|----------|

| feature, bug, spike | Worktree | --in-place |

| chore, docs | In-place | --worktree |

```bash

# Default: smart based on task type

/pokayokay:work

# Force worktree for a chore

/pokayokay:work --worktree

# Force in-place for a feature

/pokayokay:work --in-place

```

Story-based reuse: Tasks in the same story share a worktree, keeping related changes together.

On completion: Choose to merge, create PR, keep worktree, or discard work.

Sub-Agents

pokayokay includes 10 specialized sub-agents that run in isolated context windows for verbose operations:

| Agent | Model | Purpose |

|-------|-------|---------|

| yokay-auditor | Sonnet | L0-L5 completeness scanning (read-only) |

| yokay-brainstormer | Sonnet | Refines ambiguous tasks into clear requirements |

| yokay-explorer |