convez.ai
MCP ServerProdutividadeAvançado
STiFLeR7

STiFLeR's Claude Code Plugins

STiFLeR7/claude-plugins

Memória persistente e refinamento de prompts para agentes Claude

Ver no GitHub
InicialPluginScore · de 0 a 100

Marketplace de plugins para Claude focado em memória persistente de agentes e refinamento de prompts. Inclui memex-mcp para construir grafos de conhecimento temporal da codebase e prompt-forge para extrair intenção real e aplicar lentes de análise (segurança, QA, performance) antes da execução.

Instalação

$npx convez-ai STiFLeR7/claude-plugins

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
0
Forks
0
Acompanhando
0
Saúde
42%
Último commit
há 3 dias
Versão
—
Issues abertas
0
Contribuidores
2

Casos de uso

  • Manter contexto arquitetural entre sessões do Claude
  • Rastrear decisões e impactos em grandes repositórios
  • Refinar prompts com múltiplas perspectivas antes de executar
  • Reduzir cold starts em sessões de coding com agentes

Veredito

Stack técnico robusto (Neo4j + Graphiti + Gemini Flash) com dois plugins bem-definidos: memex resolve o problema real de memória em agentes IA via grafos temporais bitemporal; prompt-forge é genuinamente útil para decompor intenção complexa. MIT licensed e mantido ativo — recomendado para equipes que rodam agentes recorrentes.

README

A tradução para o português deste README ainda está na fila.

STiFLeR's Claude Code Plugins

GitHub stars memex-mcp on npm memex-mcp on PyPI License: MIT

A small, opinionated marketplace of Claude Code plugins focused on making agents remember and understand intent.

/plugin marketplace add STiFLeR7/claude-plugins
/plugin install memex-mcp@stifler-marketplace
/plugin install prompt-forge@stifler-marketplace

Plugins

🧠 memex-mcp

Persistent memory for AI coding agents. Builds a temporal knowledge graph of your codebase and serves it through 14 MCP tools — get_project_context, record_decision, predict_impact, explain_change, and more.

Every Claude session starts knowing your architecture, your decisions, and your open problems. No more cold starts. No more context pasting.

  • Bitemporal facts (every edge knows when it was true)
  • Two-regime confidence decay (unvalidated facts cross stale at exactly 30 days)
  • Hierarchical clusters (briefing stays under 1500 tokens regardless of repo size)
  • Human-in-the-loop validation via memex review
  • Anthropic memory-tool backend included

Backed by Neo4j + Graphiti + Gemini Flash. MIT licensed. → github.com/STiFLeR7/memex

✍️ prompt-forge

Advanced prompt refinement skill. Extracts your real intent, investigates your codebase and ecosystem, applies 9 perspective lenses (Security, QA, UX, Performance, …), and produces a grounded prompt for your execution tool — without ever executing the task itself.

Trigger phrases: "help me prompt", "improve my prompt", "how to write a prompt for…". The skill investigates, asks sharp clarifying questions, then hands you a refined prompt to paste into your next agent.

MIT licensed. → github.com/STiFLeR7/prompt-forge

Local development

# Clone and link as a local marketplace for testing
git clone https://github.com/STiFLeR7/claude-plugins
/plugin marketplace add /absolute/path/to/claude-plugins

Contributing

Plugins added to this marketplace should be:

  • Single-purpose and well-tested
  • Documented with clear trigger conditions
  • MIT (or compatible) licensed
  • Maintained by an active author

Open an issue with a proposal before submitting a PR.

License

MIT — see individual plugin directories for their respective licenses.

Ver o original no GitHub

Bom para

Relacionados