Spec-driven development harness for Claude Code -- 20 commands, five-phase model, five agents, three-tier domain skills, crystallize, constitutional principles. TypeScript orchestration via prunejuice MCP server.
npx claudepluginhub lewdwig-v/unslop --plugin unslopMerge file specs into a unit spec (spec granularity consolidation)
Run the adversarial quality pipeline on a managed file. Extracts behaviour, generates black-box tests, and validates via mutation.
Record a change request for a managed file
Check cross-spec coherence across related specs, including blocked-by constraint tracking
Grow test coverage on a managed file or unit using mutation-driven discovery
Extract cross-cutting patterns from the spec corpus into project-local domain skills
Infer a spec from existing code -- Archaeologist reads code and produces a candidate spec
Structured spec elicitation -- Socratic dialogue for creating or amending specs
Partition a unit spec into file specs (spec granularity splitting)
Regenerate all stale managed files from their specs
Render a Mermaid dependency graph of specs, concrete specs, and managed files
Review a spec for completeness and suggest tightening. Use --promote to promote a concrete spec to permanent.
Initialize unslop in the current project
Promote a concrete spec (.impl.md) from ephemeral to permanent
Create or edit a spec for a source file or directory
List managed files and their staleness status
Regenerate managed files from their specs
Bring existing code under spec-driven management
Run Saboteur verification synchronously on a managed file or unit
Detect intent drift between specs and code. Surfaces meaningful discrepancies and offers per-finding remediation.
Adversarial Quality framework — three-agent pipeline (Archaeologist → Mason → Saboteur) for mutation-gated test generation
Use when generating, reviewing, or promoting concrete specs — the implementation strategy layer between abstract specs (intent) and generated code. Activates during Stage B.1 of generation, during /unslop:harden --promote, and during takeover raising.
Use when generating or regenerating code from unslop spec files. Activates during /unslop:generate, /unslop:sync, and the generation step of /unslop:takeover.
Use when writing, drafting, reviewing, or editing unslop spec files (*.spec.md). Activates for spec creation, takeover spec drafting, and spec editing guidance.
Use when running the unslop takeover pipeline to bring existing code under spec management. Covers discovery, spec drafting, archiving, generation, and the convergence validation loop.
Use when the user wants to change, fix, refactor, add, or review code in a project managed by unslop (has a .unslop/ directory). Activates for any intent that would modify source files, ask about code quality, or plan structural changes. Routes the user to the correct unslop command based on their intent.
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.
Battle-tested Claude Code plugin for engineering teams — 38 agents, 156 skills, 72 legacy command shims, production-ready hooks, and selective install workflows evolved through continuous real-world use
Comprehensive PR review agents specializing in comments, tests, error handling, type design, code quality, and code simplification
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.
Tools to maintain and improve CLAUDE.md files - audit quality, capture session learnings, and keep project memory current.
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.