dspy-skills MCP — Cas d'usage, Installer & Démo en direct

Pourquoi l'utiliser

Fonctionnalités clés

DSPy signatures / modules / predictors encoded as patterns
Teleprompter (BootstrapFewShot, MIPRO) usage guidance
Evaluation + metric design
RAG pipeline scaffolds with DSPy retrievers
Opinionated integration with common LLM providers

Démo en direct

Aperçu en pratique

dspy-skill.replay ▶ prêt

0/0

Installer

Choisissez votre client

~/Library/Application Support/Claude/claude_desktop_config.json · Windows: %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "dspy-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/OmidZamani/dspy-skills",
        "~/.claude/skills/dspy-skills"
      ],
      "_inferred": true
    }
  }
}

Ouvrez Claude Desktop → Settings → Developer → Edit Config. Redémarrez après avoir enregistré.

~/.cursor/mcp.json · .cursor/mcp.json

{
  "mcpServers": {
    "dspy-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/OmidZamani/dspy-skills",
        "~/.claude/skills/dspy-skills"
      ],
      "_inferred": true
    }
  }
}

Cursor utilise le même schéma mcpServers que Claude Desktop. La config projet l'emporte sur la globale.

VS Code → Cline → MCP Servers → Edit

{
  "mcpServers": {
    "dspy-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/OmidZamani/dspy-skills",
        "~/.claude/skills/dspy-skills"
      ],
      "_inferred": true
    }
  }
}

Cliquez sur l'icône MCP Servers dans la barre latérale Cline, puis "Edit Configuration".

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "dspy-skill": {
      "command": "git",
      "args": [
        "clone",
        "https://github.com/OmidZamani/dspy-skills",
        "~/.claude/skills/dspy-skills"
      ],
      "_inferred": true
    }
  }
}

Même format que Claude Desktop. Redémarrez Windsurf pour appliquer.

~/.continue/config.json

{
  "mcpServers": [
    {
      "name": "dspy-skill",
      "command": "git",
      "args": [
        "clone",
        "https://github.com/OmidZamani/dspy-skills",
        "~/.claude/skills/dspy-skills"
      ]
    }
  ]
}

Continue utilise un tableau d'objets serveur plutôt qu'une map.

~/.config/zed/settings.json

{
  "context_servers": {
    "dspy-skill": {
      "command": {
        "path": "git",
        "args": [
          "clone",
          "https://github.com/OmidZamani/dspy-skills",
          "~/.claude/skills/dspy-skills"
        ]
      }
    }
  }
}

Ajoutez dans context_servers. Zed recharge à chaud à la sauvegarde.

claude mcp add dspy-skill -- git clone https://github.com/OmidZamani/dspy-skills ~/.claude/skills/dspy-skills

Une seule ligne. Vérifiez avec claude mcp list. Supprimez avec claude mcp remove.

Cas d'usage

Usages concrets : dspy-skills

How to build your first DSPy program and optimize it

👤 ML engineers and applied researchers ⏱ ~90 min advanced

Quand l'utiliser : You have a task where prompt quality matters and want a systematic way to improve it.

Prérequis

Python 3.10+ with dspy-ai installed — pip install dspy-ai
Skill cloned — git clone https://github.com/OmidZamani/dspy-skills ~/.claude/skills/dspy-skills

Déroulement

Define the signature

Task: classify support tickets into {billing, technical, account}. Give me a DSPy signature and a simple Predict module.✓ Copié

→ Signature + module code
Write an eval

Add an evaluation set of 50 labeled examples and an accuracy metric.✓ Copié

→ Eval harness with metric callable
Optimize

Run BootstrapFewShot to compile the module against the eval set.✓ Copié

→ Compiled predictor + improved score

Résultat : A DSPy-optimized predictor that beats a hand-written prompt, with reproducible code.

Pièges

Eval too small — optimizer overfits — Minimum 100–200 examples; hold out a true test set
Metric doesn't capture what you care about — Spend on metric design before on model choice

Combiner avec : filesystem

Build a RAG pipeline with DSPy

👤 Engineers building retrieval-augmented systems ⏱ ~120 min advanced

Quand l'utiliser : You want modular, optimizable RAG rather than a hand-wired chain.

Déroulement

Define the modules

Create a DSPy RAG pipeline: RetrieveThenRead with ColBERTv2 or a local retriever.✓ Copié

→ Modular pipeline with separate retrieval and generation
Optimize end-to-end

Write an eval on our QA set and run MIPRO to improve.✓ Copié

→ Compiled pipeline with score delta

Résultat : A RAG pipeline you can improve by changing the eval, not by rewriting prompts.

Pièges

Retriever quality caps end-to-end quality — Evaluate retrieval separately (recall@k) before optimizing generation

Combiner avec : local-rag

Combinaisons

Associez-le à d'autres MCPs pour un effet X10

dspy-skill + local-rag

Plug a local retriever into DSPy's RAG modules

Swap the ColBERT retriever for a local-rag MCP as the retrieval source.✓ Copié

dspy-skill + filesystem

Organize DSPy programs, evals, and artifacts in a repo

Lay out a DSPy project with programs/, evals/, and artifacts/ directories.✓ Copié

Outils

Ce que ce MCP expose

Outil	Entrées	Quand appeler	Coût
signature-design	task spec	Start of any DSPy program	0
module-authoring	signatures + flow	After signatures	0
teleprompter-optimization	module + eval	After eval is ready	LLM tokens during optimization
evaluation-harness	task data	Before optimizing	0

Coût et limites

Coût d'exécution

Quota d'API: LLM tokens dominate during optimization runs
Tokens par appel: Can be high — optimizations may invoke the LLM hundreds of times
Monétaire: depends on provider
Astuce: Use cheap models during teleprompter runs, upgrade only for final evaluation

Sécurité

Permissions, secrets, portée

Stockage des identifiants : LLM provider keys in env vars

Sortie de données : LLM provider endpoints

Dépannage

Erreurs courantes et correctifs

Teleprompter seems to make things worse

Check metric correctness; use a held-out set; widen the example pool.

Optimization eats your budget

Cap max_bootstrapped_demos and use cheaper models during search.

Alternatives

dspy-skills vs autres

Alternative	Quand l'utiliser	Compromis
prompt-architect-skill	You want prompt-level craft, not DSPy's programmatic approach	Hand-crafted vs optimized

Plus

Ressources

📖 Lire le README officiel sur GitHub

🐙 Voir les issues ouvertes

🔍 Parcourir les 400+ serveurs MCP et Skills