Help us improve
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
Spec-driven development with smart compaction. A Claude Code plugin that combines the Ralph Wiggum agentic loop with structured specification workflow.
npx claudepluginhub informatico-madrid/ralph-harness --plugin ralphharnessCancel active execution loop, cleanup state, and remove spec
Generate technical design from requirements
Submit feedback or report an issue for RalphHarness plugin.
Show help for RalphHarness plugin commands and workflow.
You are starting the task execution loop.
This agent should be used to "create technical design", "define architecture", "design components", "create design.md", "analyze trade-offs". Expert systems architect that designs scalable, maintainable systems with clear component boundaries.
Parallel review agent that evaluates completed tasks via filesystem communication
This agent should be used to "generate requirements", "write user stories", "define acceptance criteria", "create requirements.md", "gather product requirements". Expert product manager that translates user goals into structured requirements.
This agent should be used to "run verification task", "check quality gate", "verify acceptance criteria", "run [VERIFY] task", "execute quality checkpoint", "story verification", "exploratory verification". QA engineer that runs verification commands and outputs VERIFICATION_PASS, VERIFICATION_FAIL, or VERIFICATION_DEGRADED.
This agent should be used to "update spec files", "refactor requirements", "revise design", "modify tasks after execution", "incrementally update specifications". Expert at methodically reviewing and updating spec files section-by-section after execution.
This skill should be used when generating spec artifacts (research.md, requirements.md, design.md, tasks.md), formatting agent output, structuring phase results, or when any Ralph agent needs guidance on concise, scannable output formatting. Applies to all Ralph spec phase agents.
This skill must ALWAYS be invoked at the start of every spec — it audits the agent's own system prompt for broken references before any work begins. Invoke unconditionally regardless of goal keywords. Detects phantom infrastructure, ghost paths, incorrect URLs, missing CLI tools, and absent .env files referenced in CLAUDE.md, copilot-instructions.md, or any active system prompt instructions.
Load this skill suite for any spec that involves end-to-end testing, browser automation, Playwright, MCP browser tools, VE tasks, user flow verification, UI testing, or integration tests that drive a real browser. Covers session lifecycle, environment setup, navigation anti-patterns, stable-state detection, auth flows, selector stability, and cleanup guarantees.
This skill should be used when running an interactive interview before a spec phase, gathering requirements through dialogue, asking the user clarifying questions before delegating to a subagent, or when any Ralph phase command (research, requirements, design, tasks) needs adaptive brainstorming dialogue. Covers the 3-phase algorithm (Understand, Propose Approaches, Confirm and Store).
This skill should be used when the user asks about "ralph arguments", "quick mode", "commit spec", "max iterations", "ralph state file", "execution modes", "ralph loop", "coordinator behavior", "delegate to subagent", or needs guidance on Ralph plugin arguments, state management, delegation patterns, or execution loop behavior. Core behavioral skill for all RalphHarness operations.
Uses power tools
Uses Bash, Write, or Edit tools
Share bugs, ideas, or general feedback.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Battle-tested Claude Code plugin for engineering teams — 60 agents, 229 skills, 75 legacy command shims, production-ready hooks, and selective install workflows evolved through continuous real-world use
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Meta-prompting and spec-driven development system for Claude Code. Productivity framework for structured AI-assisted development.
Access thousands of AI prompts and skills directly in your AI coding assistant. Search prompts, discover skills, save your own, and improve prompts with AI.
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
Develop, test, build, and deploy Godot 4.x games with Claude Code. Includes GdUnit4 testing, web/desktop exports, CI/CD pipelines, and deployment to Vercel/GitHub Pages/itch.io.
BMAD structural mapper — converts BMAD PRD, epics, and architecture artifacts into RalphHarness spec files
BMAD structural mapper — converts BMAD PRD, epics, and architecture artifacts into smart-ralph spec files
Spec-driven development with smart compaction. A Claude Code plugin that combines the Ralph Wiggum agentic loop with structured specification workflow.
/ralph-bmad:import# Add the marketplace
/plugin marketplace add informatico-madrid/ralphharness
# Install the plugin
/plugin install ralphharness@ralphharness
# Restart Claude Code to load
# Clone the repo
git clone https://github.com/informatico-madrid/ralphharness.git
# Install from local path
/plugin install /path/to/ralphharness
# Or install directly from GitHub
/plugin install https://github.com/informatico-madrid/ralphharness
# Clone and link for development
git clone https://github.com/informatico-madrid/ralphharness.git
cd ralphharness
/plugin install .
When installed via the Codex-packaged distribution (ralphharness-codex), commands are exposed with the ralphharness- prefix:
$ralphharness-triage "Build a multi-tenant SaaS platform"
$ralphharness-research
$ralphharness-requirements
$ralphharness-design
$ralphharness-tasks
$ralphharness-implement
$ralphharness-start my-feature "Build user authentication"
$ralphharness-cancel
$ralphharness-status
$ralphharness-feedback
$ralphharness-help
$ralphharness-index
$ralphharness-refactor
$ralphharness-rollback
$ralphharness-switch
See the Codex plugin README for full Codex-specific documentation.
/ralphharness "Add user authentication with JWT tokens" --mode interactive --dir ./auth-spec
This will:
requirements.md and pause for approval/ralphharness:approve, generate design.md and pausetasks.md and pause# The smart way (auto-detects resume or new)
/ralphharness:start user-auth Add JWT authentication
# Quick mode (skip spec phases, auto-generate everything)
/ralphharness:start "Add user auth" --quick
# The step-by-step way
/ralphharness:new user-auth Add JWT authentication
/ralphharness:requirements
/ralphharness:design
/ralphharness:tasks
/ralphharness:implement
| Command | Description |
|---|---|
/ralphharness "goal" [options] | Start the spec-driven loop |
/ralphharness:approve | Approve current phase (interactive mode) |
/ralphharness:cancel | Cancel active loop and cleanup |
/ralphharness:feedback | Collect and process user feedback |
/ralphharness:help | Show help |
/ralphharness:index | Index/rebuild spec directory |
/ralphharness:refactor | Refactor existing spec |
/ralphharness:rollback | Rollback to git checkpoint |
/ralphharness:switch | Switch to another spec |
flowchart TB
subgraph Input
G[Goal Description]
end
subgraph Spec["Specification Phases"]
R[Requirements]
D[Design]
T[Tasks]
end
subgraph Exec["Execution Phase"]
E1[Task 1]
E2[Task 2]
EN[Task N]
PR[Parallel Reviewer<br/>external-agent<br/>task_review.md + chat.md]
end
subgraph Output
C[Complete]
end
G --> R
R -->|compact| D
D -->|compact| T
T -->|compact| E1
E1 -->|compact| E2
E2 -->|compact| EN
subgraph Review["Parallel Review (real-time)"]
PR -.->|writes to| chat.md[task_review.md]
PR -.->|reviews each task| E1
PR -.->|reviews each task| E2
PR -.->|reviews each task| EN
end
EN --> C
PR -.->|blocks on FAIL| E1
R -.->|interactive| A1{Approve?}
D -.->|interactive| A2{Approve?}
T -.->|interactive| A3{Approve?}
A1 -->|yes| D
A2 -->|yes| T
A3 -->|yes| E1
flowchart LR
subgraph Files["Persistent State"]
P[".ralph-progress.md<br/>Learnings & Progress"]
S[".ralph-state.json<br/>Loop State"]
end
subgraph Compaction
CM[Context Window<br/>Management]
end
P -->|survives| CM
S -->|tracks| CM
CM -->|preserves key context| P