S Skillteca
Receber por email
Explorar Categorias Coleções Newsletter
← Voltar para o catálogo

afm

Pesquisa e Web Autor: scouzi1966

Manter e estender o AFM (maclocal-api), um servidor LLM local compatível com OpenAI em Swift e CLI para Apple Foundation Models, modelos MLX, proxy de gateway de API e OCR de Visão. Ele suporta comandos CLI do AFM, comportamento da API OpenAI, streaming, chamada de ferramentas e saídas estruturadas.

328estrelas
Atualizado há 2 meses
Ver no GitHub ↗ Licença: MIT

Como adicionar

/plugin marketplace add scouzi1966/maclocal-api

O comando exato pode variar conforme o repositório. Confira o README no GitHub.

Para o autor da skill

Cole no README do seu repo

Mostra que sua skill está catalogada na Skillteca, gera backlink e tráfego rastreável.

[![Skillteca](https://www.skillteca.com.br/api/badge/afm/svg)](https://www.skillteca.com.br/skills/afm)
#ai #api #llm #ocr

Skills relacionadas

Ver todas de Pesquisa e Web →

understand-domain

64.4k

Extrai conhecimento do domínio de negócio de uma base de código e gera um grafo interativo de fluxo de domínio. Funciona de forma autônoma (varredura leve) ou deriva de um grafo de conhecimento /understand existente.

Pesquisa e Web #ai por Lum1104

understand-dashboard

64.4k 2

Lance o painel web interativo para visualizar o grafo de conhecimento de uma base de código.

Pesquisa e Web por Lum1104

understand-chat

64.4k

Use quando precisar fazer perguntas sobre uma base de código ou entender código usando um grafo de conhecimento.

Pesquisa e Web por Lum1104

dev-browser

63k

Automatiza interações do navegador com estado de página persistente. Utilize para navegar em sites, preencher formulários, tirar capturas de tela, extrair dados da web, testar aplicativos ou automatizar fluxos de trabalho do navegador.

Pesquisa e Web #test por code-yeongyu

Alerta por categoria

Receba novas skills de Pesquisa e Web toda segunda

1 email curto, só com as skills novas de Pesquisa e Web. 4 minutos de leitura, sem spam, cancela com 1 clique.

Você confirma o email no primeiro envio. Sem spam. Cancela com 1 clique.

Ler descrição completa

AFM Skill

AFM (maclocal-api) is a local OpenAI-compatible API server and CLI for:

  • Apple Foundation Models (afm)
  • MLX models from Hugging Face (afm mlx)
  • Local backend gateway/proxy mode (afm -g)
  • Vision OCR/table extraction (afm vision)

1. Overview and Triggers

Use this skill when working on:

  • CLI flags, argument dispatch, or mode routing (afm, afm mlx, afm vision)
  • OpenAI-compatible routes (/v1/chat/completions, /v1/models)
  • Streaming SSE behavior and response compatibility
  • MLX tool calling, parser overrides, reasoning extraction, structured outputs
  • Gateway backend discovery/proxying (Ollama, LM Studio, Jan, local OpenAI endpoints)
  • Vision OCR and table extraction
  • Vendor mlx-swift-lm patch pipeline, build scripts, regression tests

2. Key File Reference

| Purpose | File Path | |---------|-----------| | CLI modes, flags, single-prompt and stdin behavior | Sources/MacLocalAPI/main.swift | | Vision command | Sources/MacLocalAPI/VisionCommand.swift | | Vapor server + route registration | Sources/MacLocalAPI/Server.swift | | Foundation chat completions controller | Sources/MacLocalAPI/Controllers/ChatCompletionsController.swift | | MLX chat completions controller | Sources/MacLocalAPI/Controllers/MLXChatCompletionsController.swift | | Foundation model service | Sources/MacLocalAPI/Models/FoundationModelService.swift | | MLX model service | Sources/MacLocalAPI/Models/MLXModelService.swift | | OpenAI request schema | Sources/MacLocalAPI/Models/OpenAIRequest.swift | | OpenAI response schema | Sources/MacLocalAPI/Models/OpenAIResponse.swift | | Backend discovery | Sources/MacLocalAPI/Services/BackendDiscoveryService.swift | | Backend proxying | Sources/MacLocalAPI/Services/BackendProxyService.swift | | Backend definitions/capabilities | Sources/MacLocalAPI/Models/BackendConfiguration.swift | | Vendor patch orchestrator | Scripts/apply-mlx-patches.sh | | Full bootstrap build | Scripts/build-from-scratch.sh |

3. Core Rules

  • Work from maclocal-api/ project root.
  • Keep OpenAI compatibility first: request fields, response shape, SSE chunk format.
  • Keep Foundation and MLX behavior aligned where practical.
  • Treat vendor/mlx-swift-lm as patch-target output:
  • edit Scripts/patches/* first
  • apply/check via Scripts/apply-mlx-patches.sh
  • Preserve CORS and streaming headers.
  • Keep both streaming and non-streaming paths correct for every new feature.

4. Quick Start Commands

# Build (Makefile applies vendor patches)
make build

# Debug build
make debug

# Full bootstrap build
./Scripts/build-from-scratch.sh

# Vendor patch workflow
./Scripts/apply-mlx-patches.sh --check
./Scripts/apply-mlx-patches.sh

# Run Foundation server
./.build/debug/afm -p 9999 -v

# Run MLX server
./.build/debug/afm mlx -m mlx-community/Qwen3-0.6B-4bit -p 9999 -v

# Single prompt
./.build/debug/afm mlx -m mlx-community/Qwen3-0.6B-4bit -s "hello"

# Vision OCR
./.build/debug/afm vision -f media/ocr.png

5. Primary Workflow

  1. Identify target mode and load only relevant reference files.
  2. Update shared request/response models first when adding API-visible behavior.
  3. Thread parameters through CLI -> server/controller -> service -> response.
  4. Validate non-streaming and streaming paths.
  5. Run targeted checks, then broader regression scripts.
  6. Update README examples if user-visible behavior changed.

6. Reference Links

| Reference | When to Use | |-----------|-------------| | references/architecture.md | Route wiring, lifecycle, cross-mode request flow | | references/cli-and-modes.md | Flag behavior, command dispatch, mode defaults | | references/api-contract.md | OpenAI request/response compatibility details | | references/mlx-inference-and-tool-calling.md | MLX streaming, tool calls, reasoning extraction, multimodal | | references/gateway-discovery-and-proxy.md | Gateway scans, backend capabilities, proxy normalization | | references/patch-build-test.md | Vendor patch process, build scripts, regression commands |

7. Validation Checklist

  • CLI parse/help:
  • ./.build/debug/afm --help
  • ./.build/debug/afm mlx --help
  • ./.build/debug/afm vision --help
  • Server health/models:
  • curl http://127.0.0.1:9999/health
  • curl http://127.0.0.1:9999/v1/models
  • Regression scripts:
  • ./test-all-features.sh
  • ./Scripts/afm-cli-tests.sh
  • ./test-streaming.sh
  • ./test-go.sh
  • ./test-metrics.sh
Compartilhar X LinkedIn