opik-mcp MCP — Casos de uso, Instalar & Demo en vivo

Por qué usarlo

Características clave

Official, maintained by Comet
Prompt lifecycle: list, get, version, promote
Workspace / project / trace exploration
Dataset + metric operations for evals
Cloud (comet.com) or self-hosted; HTTP bearer auth supported

Demo en vivo

Cómo se ve en la práctica

opik.replay ▶ listo

0/0

Instalar

Elige tu cliente

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

{
  "mcpServers": {
    "opik": {
      "command": "npx",
      "args": [
        "-y",
        "opik-mcp"
      ],
      "_inferred": true
    }
  }
}

Abre Claude Desktop → Settings → Developer → Edit Config. Reinicia después de guardar.

~/.cursor/mcp.json · .cursor/mcp.json

{
  "mcpServers": {
    "opik": {
      "command": "npx",
      "args": [
        "-y",
        "opik-mcp"
      ],
      "_inferred": true
    }
  }
}

Cursor usa el mismo esquema mcpServers que Claude Desktop. La configuración del proyecto prevalece sobre la global.

VS Code → Cline → MCP Servers → Edit

{
  "mcpServers": {
    "opik": {
      "command": "npx",
      "args": [
        "-y",
        "opik-mcp"
      ],
      "_inferred": true
    }
  }
}

Haz clic en el icono MCP Servers de la barra lateral de Cline y luego en "Edit Configuration".

~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "opik": {
      "command": "npx",
      "args": [
        "-y",
        "opik-mcp"
      ],
      "_inferred": true
    }
  }
}

Mismo formato que Claude Desktop. Reinicia Windsurf para aplicar.

~/.continue/config.json

{
  "mcpServers": [
    {
      "name": "opik",
      "command": "npx",
      "args": [
        "-y",
        "opik-mcp"
      ]
    }
  ]
}

Continue usa un array de objetos de servidor en lugar de un mapa.

~/.config/zed/settings.json

{
  "context_servers": {
    "opik": {
      "command": {
        "path": "npx",
        "args": [
          "-y",
          "opik-mcp"
        ]
      }
    }
  }
}

Añádelo a context_servers. Zed recarga en caliente al guardar.

claude mcp add opik -- npx -y opik-mcp

Un solo comando. Verifica con claude mcp list. Quita con claude mcp remove.

Casos de uso

Usos del mundo real: opik-mcp

Pull a production trace into your IDE to debug a bad LLM response

👤 LLM app developers ⏱ ~15 min intermediate

Cuándo usarlo: A user reports a wrong answer; the trace is in Opik; you want to inspect it without leaving Cursor.

Requisitos previos

Opik API key — comet.com/site > API Keys (or self-hosted admin)

Flujo

Find the trace

Search traces in project 'prod-chatbot' where output contains 'I cannot help with that'. Last 24h.✓ Copiado

→ Matching trace IDs + timestamps
Inspect

Open trace ID abc123. Show me the full message chain, tools called, and intermediate reasoning.✓ Copiado

→ Full trace object
Form hypothesis

Why might the model have refused? Compare this trace to a successful one on the same prompt template.✓ Copiado

→ Diff + hypothesis

Resultado: Faster trace-driven debugging without app-switching.

Errores comunes

PII in traces — Configure Opik's redaction before enabling MCP access broadly

Iterate on a prompt template with version tracking

👤 Prompt engineers ⏱ ~25 min advanced

Cuándo usarlo: You're tuning a system prompt and want each version saved to Opik for rollback.

Flujo

Pull current version

Get latest version of prompt 'support-agent-system'.✓ Copiado

→ Current prompt body
Edit and commit

Propose a change to handle escalations better. Show diff. Commit as a new version with message 'add escalation path'.✓ Copiado

→ Diff + new version ID
Eval against dataset

Run this new version against dataset 'support-eval-v1'. Compare pass rate vs previous version.✓ Copiado

→ Metric comparison

Resultado: Data-driven prompt changes, version-controlled.

Errores comunes

No guardrails — a regressive prompt becomes prod — Use Opik's experiment framework: don't promote until pass rate ≥ baseline

Generate a weekly LLM app health report

👤 Eng leads, LLM app PMs ⏱ ~30 min intermediate

Cuándo usarlo: You want a Monday-morning digest of cost, latency, error rate, and top failure categories.

Flujo

Pull last week's metrics

For project 'prod-chatbot': total traces, total tokens, avg latency p50/p95, error count — over last 7 days.✓ Copiado

→ Metrics block
Classify failures

Sample 20 failed traces. Cluster by failure mode. Rank clusters by frequency.✓ Copiado

→ Failure taxonomy
Write the digest

Compose a Markdown digest with the metrics and top 3 failure modes, ready for Slack.✓ Copiado

→ Shareable report

Resultado: Weekly LLM ops awareness without manual dashboard time.

Errores comunes

Metric drift as your app evolves — Version the report template; compare apples to apples week over week

Combinar con: notion

Combinaciones

Combínalo con otros MCPs para multiplicar por 10

opik + github

When a prompt regresses, open a GitHub issue with the failing trace

If pass rate drops >5% on 'support-eval-v1' vs last week, create a GitHub issue with the top 3 failing trace IDs.✓ Copiado

opik + notion

Publish weekly LLM health digest to Notion

Compose a Monday digest from last week's Opik metrics and create a Notion page in 'LLM Weekly'.✓ Copiado

Herramientas

Lo que expone este MCP

Herramienta	Entradas	Cuándo llamar	Coste
list_projects	workspace_id?	Navigate your workspace	1 API call
list_traces	project, filter?, start?, end?, limit?	Find traces by time range or content	1 API call
get_trace	trace_id	Deep-dive a single trace	1 API call
get_prompt	name, version?	Read a prompt for editing or use in code	1 API call
create_prompt_version	name, template, message?	Commit a new prompt iteration	1 API call
create_dataset	name, items[]	Build an eval dataset	1 API call
get_metrics	project, metric, window	Monitor cost / latency / quality	1 API call

Coste y límites

Lo que cuesta ejecutarlo

Cuota de API: Opik Cloud has per-plan limits; self-hosted is unlimited
Tokens por llamada: Trace listings 1k-5k tokens; single traces 500-3000
Monetario: Opik has a generous free tier; paid plans for scale. MCP itself is free (Apache 2.0).
Consejo: Use list_traces with a time window; never call without a range on a busy project.

Seguridad

Permisos, secretos, alcance

Ámbitos mínimos: Opik API key scope the workspace you intend to expose

Almacenamiento de credenciales: OPIK_API_KEY env var; HTTP transport uses Authorization: Bearer

Salida de datos: Traces may contain prompts/responses with PII — understand your Opik region and redaction setup

No conceder nunca: An admin-scope key to a shared dev machine

Trace data is often PII-heavy (user prompts). Redact upstream before Opik ingests.
For self-hosted, the MCP needs network reach to your Opik backend — include it in your VPN/firewall plan.

Resolución de problemas

Errores comunes y soluciones

401 Unauthorized (Bearer)

Check OPIK_API_KEY. For self-hosted, also set --apiUrl http://host:5173/api.

Verificar: curl -H 'Authorization: Bearer $KEY' $URL/api/v1/workspaces

Empty trace list despite traffic

Wrong project / workspace. List projects first and confirm UUID.

Self-hosted MCP can't reach backend

Use container networking (same docker network) or map --apiUrl to an externally-reachable URL.

Alternativas

opik-mcp vs otros

Alternativa	Cuándo usarla	Contrapartida
LangSmith MCP	You use LangSmith for tracing	Different platform; similar capabilities
Langfuse MCP	You use Langfuse (OSS)	Also OSS + self-hostable; different schemas
Arize / Phoenix	You want focus on evals + drift detection	Richer ML-monitoring features; steeper learning curve

Más

Recursos

📖 Lee el README oficial en GitHub

🐙 Ver issues abiertas

🔍 Ver todos los 400+ servidores MCP y Skills