Skip to content

flashwade03/fablers-claude-plugins

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

44 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

fablers

Stop prompting. Start harnessing.

A Claude Code plugin marketplace packaging battle-tested workflows into reusable plugins. Design methodology, document forging, and agentic RAG — all in one place.

Claude Code Marketplace Version License: MIT

English | 한국어 | 日本語


Plugins

vibe-architecture — Design methodology & skill extraction

Encodes proven design methodologies into skills that Claude loads on demand. No more ad-hoc prompting — just structured workflows that actually work.

Skill What it does
vibe-design Turn rough ideas into just-enough design specs. Decisions + Constraints + Milestones, never pseudocode.
design-review Score design docs against 6 axes (S~F grade, 0-100 score). Any FAIL caps the grade at C.
sketch-team Agent Teams workflow for concrete multi-domain design: 1–3 Specialist Designers (data-model / api-surface / protocol / etc., Lead-decided) produce concrete artifacts with inline rejected alternatives, Planner composes with cross-domain coherence checks, 2 Reviewers judge against a concretion-friendly 6-axis rubric, loop until approved.
session-skill-extractor Analyze conversations to extract reusable patterns. Routes findings to skills, CLAUDE.md, hookify, or memory.
> /sketch                          # Start a design (single-agent Q&A)
> /sketch-team <task>              # Agent Teams: design + review bundled, auto-iterates
> 설계 리뷰해줘                      # Review a design
> 대화에서 스킬 추출해줘              # Extract skills from session

damascus — Forge documents through iterative multi-LLM review

Like Damascus steel, documents become stronger through repeated forging.

Refines documents through an iterative review loop powered by multiple LLMs. Write an implementation plan or technical document, have it reviewed by Claude, Gemini, and OpenAI in parallel, then refine until approved.

> /forge <task description>        # Single-reviewer forging
> /forge-team <task description>   # Multi-LLM team review
> /forge-plan <task description>   # Plan-only mode
> /forge-doc <task description>    # Document-only mode

fablers-agentic-rag — Ask your documents. Get a cited answer.

Agentic RAG pipeline — query analysis, hybrid retrieval (vector + BM25), CRAG validation, and cited answer synthesis — all orchestrated by Claude agents. Supports PDF, plain text, and Markdown.

> /ingest <file>                   # Index a document
> /rag-ask <question>              # Query with citations
> /rag-search <query>              # Raw retrieval

grimoire — General-purpose reusable skills

Not a single-purpose plugin. A growing collection of Claude Code skills that aren't tied to any one command or workflow — the kind Claude should pull out whenever the situation calls for them, regardless of which other plugin is active. Named grimoire because it's a book of reusable patterns; the description you're reading compensates for the metaphor.

Skill What it does
agent-teams Run real Claude Code agent teams (TeamCreate + SendMessage) for debates, reviews, and implementation — instead of letting a single agent role-play multiple personas. Includes cross-challenge patterns, worked examples, and explicit scope boundaries against /forge-team.
> 에이전트 팀 만들어줘               # Spawn a real agent team
> debate with agent teams          # Structured cross-agent debate

Quick Start

# Add the marketplace
/plugin marketplace add flashwade03/fablers-claude-plugins

# Install individual plugins
/plugin install vibe-architecture@fablers
/plugin install damascus@fablers
/plugin install fablers-agentic-rag@fablers
/plugin install grimoire@fablers

Project Structure

my-claude-harness/
├── .claude-plugin/
│   └── marketplace.json
├── plugins/
│   ├── vibe-architecture/         # Design methodology
│   │   ├── .claude-plugin/plugin.json
│   │   ├── commands/
│   │   └── skills/
│   │       ├── vibe-design/
│   │       ├── design-review/
│   │       └── session-skill-extractor/
│   ├── damascus/                       # Document forging
│   │   ├── .claude-plugin/plugin.json
│   │   ├── agents/
│   │   ├── commands/
│   │   ├── hooks/
│   │   ├── scripts/
│   │   └── skills/
│   ├── fablers-agentic-rag/            # Agentic RAG
│   │   ├── .claude-plugin/plugin.json
│   │   ├── agents/
│   │   ├── commands/
│   │   ├── hooks/
│   │   ├── scripts/
│   │   └── skills/
│   └── grimoire/                       # General-purpose reusable skills
│       ├── .claude-plugin/plugin.json
│       └── skills/
│           └── agent-teams/

Philosophy

  1. Decisions, not implementation — Design documents record what and why, never how
  2. Progressive disclosure — Core workflow loads first, details load only when needed
  3. Earn your complexity — Simple patterns stay simple. Skills are created only when the pattern justifies one

License

MIT

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors